{
  "meta": {
    "title": "Trade Accountability System - Todo List",
    "owner": "James",
    "description": "Single source of truth for work across sessions. Paste https://trsportsacademy.com/todos/ into any new Claude session to catch up.",
    "updated_at": "2026-08-16T23:52:00-04:00",
    "last_human_review_at": "2026-05-12T23:09:27-04:00",
    "last_human_review_summary": "Merged 11 task(s) from /todos/pending_tasks.json via merge_pending.php at 2026-05-12T23:09:27-04:00. (prior: Merged 1 task(s) from /todos/pending_tasks.json via merge_pending.php (new properties: alerts-hotkeys-etc) at 2026-05-12T07:01:19-04:00. (prior: Cleaned up 2 smoke-test/demo entries (test, smoke-test-v2-merge-pending-pipeline-delete-me) after end-to-end verification of the new add-task form + immediate-merge pipeline. Form fields validated: priority P0/P1/P2/P3 order, generic lifecycle phases (des\u2026)",
    "schema_version": 3,
    "last_nightly_run": "2026-08-16T23:05:00-04:00",
    "last_nightly_run_note": "Sunday. No backfill needed - daily reports continuous through 08-15. Pending queue empty, Step 2a skipped. THE SCAN FAILED OVER: all three WebDAV scan calls and the startup heartbeat hung past 20 minutes and were killed; the entire day's evidence came from a server-side `ssh a2 find` that returned in seconds. Raised nightly-scan-webdav-stall to P1. Shipped today: LYNCH_MASTER.html rebuilt 22:00 (third sweep, winner byte-identical to 08-09 - logged lynch-window-anchored-verify to find out whether the window is frozen), J3P_WEEK_AHEAD.html 10:00 with 24 names to prepare, and an isolated accountability-staging-entity workspace (~30 files, 20:25) giving the smb-reports-v2 accountability tree a real dev lane. Market closed: 16:45 EOD catch-up wrote an empty scorecard (0 trades, correct) and refreshed you-vs-you; the multi-trader step failed AGAIN and now fails with zero log output. you-vs-you moved from -$1,435 to -$1,564 purely because a green day rolled out of the 21->20 day window, not from new losses. Adaptive_Mean_Portfolio_V1.0.txt edited a fifth consecutive day; grep proves zero payloads under its tag all-time. Corrected the 08-14/08-15 'HUD is writing errors' claim - the error_log holds DB-connect success chatter, already tracked since 08-02; dropped to P3.",
    "last_nightly_report_date": "2026-08-16",
    "last_nightly_report_url": "/todos/daily/2026-08-16.html",
    "next_nightly_run": "2026-08-17T23:00:00-04:00",
    "today_focus": [
      {
        "task_id": "trade-count-cap-live-meter-2026-08-15",
        "title": "Build the 10-trade cap counter before the 09:30 open",
        "why": "Monday is a trading day and this is the only P0 on the board. The Simple Version says capping the day at 10 trades flips 60 days from -$2,104 to +$504 with fees falling ~$2,000 to ~$110, and tonight's you-vs-you says the same thing from a different angle: 1,515 sub-2-minute trades cost -$2,657, and removing them alone turns the period from -$1,564 to +$1,093. Two independent scoreboards, one rule. It has now sat as prose for two days. Auto-populate, not manual - a live count-down on a J3P surface; the cockpit stays read-only. If it cannot be live by 09:30, a paper tally beats nothing."
      },
      {
        "task_id": "eod-multitrader-crash-2026-08-15",
        "title": "Make the multi-trader step log its own failure",
        "why": "It broke differently tonight and got harder to see: no traceback, no 'catch-up done', the log just stops. It still wrote traders.json and one card before dying, so it is getting further than Saturday and then vanishing. Two nights running, the thing that is missing is the evidence needed to fix it - so fix the logging first, then the bug. Cheap, and it unblocks the all-trader view."
      },
      {
        "task_id": "nightly-scan-webdav-stall-2026-08-14",
        "title": "Move the nightly scan onto the ssh a2 walk",
        "why": "Tonight it stopped being slow and started being broken - nothing returned at all. The replacement is already proven: the same 19 roots over ssh came back in seconds instead of stalling past 20 minutes, and a working reference implementation is in the scratchpad. Until this lands, every nightly is one WebDAV stall away from a writeup with no evidence behind it."
      }
    ],
    "today_focus_set_at": "2026-08-16T23:52:00-04:00",
    "projects_tagged_at": "2026-06-20T16:30:00-04:00"
  },
  "current_priority_id": "acct-v2-bug-risk-bump-ignores-date-range",
  "properties": [
    {
      "id": "strategy_library",
      "label": "Strategy Library",
      "lens": "DEFINE",
      "zone": "PRE-TRADE",
      "url": "/strategies/"
    },
    {
      "id": "playbooks",
      "label": "Playbooks",
      "lens": "TEACH",
      "zone": "PRE-TRADE",
      "url": "/playbooks/"
    },
    {
      "id": "mgp",
      "label": "Morning Game Plan",
      "lens": "PLAN",
      "zone": "PRE-TRADE",
      "url": "/mgp/"
    },
    {
      "id": "accountability_hud",
      "label": "Accountability HUD",
      "lens": "EXECUTE",
      "zone": "LIVE COCKPIT",
      "url": "https://smb.a1prospect.com/live-accountability-v2/hud/",
      "dev_url": "https://smb.a1prospect.com/live-accountability-v2-dev/hud/"
    },
    {
      "id": "scorecards",
      "label": "Scorecards",
      "lens": "GRADE",
      "zone": "POST-TRADE",
      "url": "/scorecards/"
    },
    {
      "id": "monthly_reports",
      "label": "Monthly Reports",
      "lens": "ANALYZE",
      "zone": "POST-TRADE",
      "url": "https://smb.a1prospect.com/db_reports/james-monthly-reports/"
    },
    {
      "id": "param_surface",
      "label": "Param Surface",
      "lens": "OPTIMIZE",
      "zone": "POST-TRADE",
      "url": "https://smb.a1prospect.com/db_reports/smb-reports-v2/vwap-bands/param-surface/"
    },
    {
      "id": "live_classifier",
      "label": "Live Classifier HUD",
      "lens": "LIVE",
      "zone": "LIVE COCKPIT",
      "url": "/scorecards/james-live.html"
    },
    {
      "id": "daily_stop",
      "label": "Daily Stop tool",
      "lens": "LIVE",
      "zone": "LIVE COCKPIT",
      "url": "/daily-stop/"
    },
    {
      "id": "risk_hud",
      "label": "Risk Manager HUD",
      "lens": "LIVE",
      "zone": "LIVE COCKPIT",
      "url": "/accountability/risk.html"
    },
    {
      "id": "mtf_vp_scanner",
      "label": "MTF VP Scanner",
      "lens": "OPTIMIZE",
      "zone": "POST-TRADE",
      "url": "https://smb.a1prospect.com/db_reports/james-monthly-reports/mtf-vp/"
    },
    {
      "id": "foundation",
      "label": "Foundation",
      "lens": "INFRA",
      "zone": "CROSS-CUTTING",
      "url": "/system/"
    },
    {
      "id": "pm_system",
      "label": "PM System",
      "lens": "INFRA",
      "zone": "CROSS-CUTTING",
      "url": "/todos/"
    },
    {
      "id": "alerts-hotkeys-etc",
      "label": "Alerts+HotKeys+Etc",
      "lens": "DEFINE",
      "zone": "CUSTOM",
      "url": "/system/"
    },
    {
      "id": "ams",
      "label": "AMS (Accountability Management System)",
      "lens": "DEFINE",
      "zone": "PRE-TRADE",
      "url": "/ams/"
    }
  ],
  "phases": [
    {
      "id": "phase-0",
      "title": "Phase 0 - FOUNDATION PASS (FREEZE on new features until this lands)",
      "description": "One shared data layer, one freshness signal, one orchestrator, one identity fn. Kill duplicate computations. Make Param Surface a strategy registry. Every property talks to the same spine before we build anything else."
    },
    {
      "id": "phase-1",
      "title": "Phase 1 - Journal persistence",
      "description": "Stop the daily data loss. Notes from the scorecard journal must land in DB, not localStorage. (Code shipped; waiting on user steps.)"
    },
    {
      "id": "phase-2",
      "title": "Phase 2 - Webhook enrichment pipeline (THE unlock)",
      "description": "Restore indicator_alerts data flow. One pipe unlocks per-trade grading, stop/tp on charts, overtrading detection, and retires webhook-data N/A everywhere."
    },
    {
      "id": "phase-3",
      "title": "Phase 3 - Charts on scorecards",
      "description": "Per-trade thumbnails with stop/tp lines. Inherits chart invariants (ET, session-only, stacked-axis sync)."
    },
    {
      "id": "phase-4",
      "title": "Phase 4 - Multi-trader scorecards",
      "description": "Parameterize gen_scorecard.py for Enrique, Preston, Nolan, Brian, Thomas."
    },
    {
      "id": "phase-parallel",
      "title": "Parallel - Playbook chart bugs",
      "description": "Different system, fix when someone can focus on it."
    },
    {
      "id": "phase-2_5",
      "title": "Phase 2.5 - Shared user layer (unlocks every page at once)",
      "description": "One user dropdown, one identity function, role + is_approved on main.users. Prerequisite for multi-trader and strategy library."
    },
    {
      "id": "phase-5",
      "title": "Phase 5 - Strategy Library + Me-vs-Firm",
      "description": "Users upload strategy JSON, backtest against their trades, admin approve/deny. Peer comparison on Scorecards + Monthly."
    },
    {
      "id": "phase-mgp",
      "title": "Phase MGP - Morning Game Plan pipeline",
      "description": "Daily pre-market doc: pulled from SMB emails + Ned Davis breadth model. Template-driven like playbooks. Becomes the 7th property."
    },
    {
      "id": "phase-strategy-rules",
      "title": "Phase STRATEGY - Named strategy rule engines (VWAP Pull-In, Back$ide Scalp) + walk-forward",
      "description": "Fash Late is webhook-driven (SMB_FASH_EVT_*, SMB_FL2_KORE) and has a backtest at /backtests/fl2-band-trail/. VWAP Pull-In and Back$ide Scalp are Claude/KITE-built MGP-level classifications with no programmed rules yet. Each needs: rules codified, walk-forward backtest, scorecard integration, and alignment to the Strategy Breakdown pie on Scorecards."
    },
    {
      "id": "phase-project-mgmt",
      "title": "Phase PM - Nightly project-manager report + progress calendar",
      "description": "Every night at 11PM ET: scan Z drive for file changes, gather Claude conversation summaries, write daily progress report, update todos, visualize where work landed. Progress calendar view so James can see what shipped each day."
    },
    {
      "id": "phase-hold",
      "title": "Hold - lower priority",
      "description": "Do after main phases land."
    }
  ],
  "tasks": [
    {
      "id": "multi-tf-index-stack-v4-2026-06-13",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Multi-TF Index Stack v4 dev page (Bressert Fig 10.4 + Shannon 4-stage) \u2014 parallel build at /system/multi-tf-stack/",
      "path": "/z/setup_daily_reporting/reports/gen_multi_tf_stack.py",
      "created_at": "2026-06-13",
      "completed_at": "2026-06-13",
      "note": "SHIPPED 2026-06-13 (Sat). New chiphazzard_v4.py multi-timeframe stage classifier: gives each TF (1D/65m/30m/15m/5m) a state INDEPENDENT of the others, fixing the v2.2/v3 redundancy where all panels collapsed to the same daily MAs. Two framework components: (1) Bressert MA periods Fig 10.4 \u2014 SHORT/INTERMEDIATE/LONG MA proportional to each TF's bars-per-session, confirmation MA converging to daily 5DMA in TF-native bars (65\u00d730m = 5 days); (2) Brian Shannon 4-stage structure (Accumulation/Markup/Distribution/Decline, each with 3 sub-phases) + Change-of-Character header. Standalone generator gen_multi_tf_stack.py renders ONLY the new v4 panels to /system/multi-tf-stack/<date>.html (parallel-build URL, zero risk to prod v1/v2 morning/midday/recap). Supporting: gen_multi_tf_stack_tick.py (tick driver), multi_tf_stack_daemon.py (daemon), multi_tf_stack_cache/ (per-symbol 5m+1D source cache for SPY/QQQ/IWM/SMH/VIX/UVXY), market_sentiment.py. First page live for 2026-06-12 data. Extends the MTF Trend Matrix framework (mtf_trend.py, live on scorecards since 5/6) into a dedicated index-stack view.",
      "property": "mgp",
      "project": "multi_trader"
    },
    {
      "id": "mgp-v2-generator-iterations-2026-06-13",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "MGP v2 generator iterations + Dale VWAP state module",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_v2.py",
      "created_at": "2026-06-13",
      "completed_at": "2026-06-13",
      "note": "SHIPPED 2026-06-13 (Sat). Iterated gen_mgp_v2.py + gen_mgp_intraday_v2.py + gen_mgp.py/gen_mgp_intraday.py. New dale_vwap_state.py (Dale VWAP state module \u2014 1st-dev-only per project_strategy_framework convention). Pre-market universe generators refreshed (universe_cache/pvp.json, pbs.json). _shared/tooltip.js shared UI helper updated. Weekend build day \u2014 no market, all MGP editions correctly idle.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "param-surface-scorecards-refresh-2026-06-13",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Param Surface + monthly reports + scorecards weekend refresh",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "created_at": "2026-06-13",
      "completed_at": "2026-06-13",
      "note": "Regenerated 2026-06-13 (Sat). VWAP-bands Param Surface refreshed: per-symbol surface+heatmap PNGs (CROX/QBTS/BTDR/ALOY + combined), full TOD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion, param-surface-db index. james-monthly-reports june-2026 + vwap-bands index pages rebuilt. Scorecards: trade_scorecard_2026-06-13.html + manifest + you-vs-you.html regenerated.",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "phase-0-kickoff",
      "phase": "phase-0",
      "status": "completed",
      "title": "PHASE 0 KICKOFF: confirm foundation pass plan, then execute in order",
      "owner": "user",
      "priority": "P0",
      "note": "CLOSED 2026-05-08 \u00e2\u20ac\u201d Phase 0 plan confirmed and execution underway (p0-* deliverables tracked individually). The meta-confirmation that this task represented is satisfied. Original note retained: Freeze on new features until 6 foundation deliverables ship. See phase-0 tasks below.",
      "created_at": "2026-04-17",
      "property": "foundation",
      "completed_at": "2026-05-08",
      "project": "atlas"
    },
    {
      "id": "p0-freshness-dashboard",
      "phase": "phase-0",
      "status": "pending",
      "title": "1/6 Data-freshness dashboard: /system/freshness.json + banner on hub",
      "owner": "claude",
      "priority": "P0",
      "note": "STATUS 2026-05-06: schema scaffolded at /system/freshness.json AND seeded with property mtimes (9 fresh / 1 stale / 3 external). update_freshness.py written. Outstanding: cron the script + DB-source half. Bumped 6+ days; would have caught all 3 silent regressions in the last 9 days.",
      "created_at": "2026-04-17",
      "property": "foundation",
      "project": "atlas",
      "sprint": 1
    },
    {
      "id": "p0-orchestrator",
      "phase": "phase-0",
      "status": "pending",
      "title": "2/6 Pipeline orchestrator: run_nightly.py calls every generator, logs to one place, fails loud",
      "owner": "claude",
      "priority": "P0",
      "note": "Replaces cron uncertainty. Log to /z/logs/nightly-YYYY-MM-DD.log. Any step returning empty = non-zero exit + alert.",
      "created_at": "2026-04-17",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "p0-daily-summary-view",
      "phase": "phase-0",
      "status": "pending",
      "title": "3/6 main.daily_summary SQL view: kills duplicate computation",
      "owner": "claude",
      "priority": "P0",
      "note": "One view: (user_id, trade_date, pnl, trade_count, win_rate, commissions, ...). Accountability + Scorecards + Monthly all read FROM IT instead of computing themselves. Numbers match by construction.",
      "created_at": "2026-04-17",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "p0-identity-fn",
      "phase": "phase-0",
      "status": "pending",
      "title": "4/6 Identity function: resolve_user(account_id) -> (user_id, trader_name)",
      "owner": "claude",
      "priority": "P0",
      "note": "One Python helper + one PHP helper. Every generator imports. Kills DEFAULT_ACCOUNT=14859 hardcodes. Adding a trader = row in mapping table.",
      "created_at": "2026-04-17",
      "property": "foundation",
      "project": "ams"
    },
    {
      "id": "p0-param-surface-registry",
      "phase": "phase-0",
      "status": "pending",
      "title": "5/6 Param Surface as strategy registry: /param-surface/ index + per-strategy grid pages",
      "owner": "claude",
      "priority": "P0",
      "note": "Each strategy has its own grid (VWAP bands 3x4 first, others same template). Registry index lists strategies with last-refreshed + freshness badge. Storage behind it: main.user_strategies (future). Built-in strategies are rows seeded by code.",
      "created_at": "2026-04-17",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "p0-indicator-alerts-ddl",
      "phase": "phase-0",
      "status": "pending",
      "title": "6/6 indicator_alerts DDL audit: dump schema, commit as ground truth",
      "owner": "user",
      "priority": "P0",
      "note": "psql -c '\\d alerts_tradingview.indicator_alerts' > /z/trading/a1p_journal/docs/indicator_alerts_schema.txt. Needed before any 'add missing webhook fields' work.",
      "created_at": "2026-04-17",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "diag-journal",
      "phase": "phase-1",
      "status": "completed",
      "title": "Diagnose journal save path",
      "note": "Result: localStorage-only, never hit DB.",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "write-save-php",
      "phase": "phase-1",
      "status": "completed",
      "title": "Write save_journal.php endpoint (POST upsert + GET fetch)",
      "path": "/z/public_html/scorecards/api/save_journal.php",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "write-ddl-migration",
      "phase": "phase-1",
      "status": "completed",
      "title": "Write add_scorecard_columns.sql (market_conditions, emotional_state)",
      "path": "/z/trading/a1p_journal/src/ddl/add_scorecard_columns.sql",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "patch-gen-scorecard",
      "phase": "phase-1",
      "status": "completed",
      "title": "Patch gen_scorecard.py saveJournal/loadJournal to POST/GET server",
      "path": "/z/setup_daily_reporting/reports/gen_scorecard.py",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "rescue-0416",
      "phase": "phase-1",
      "status": "completed",
      "title": "Rescue 4/16 IBIT notes to durable file",
      "path": "/z/public_html/scorecards/api/_rescued_journals/2026-04-16.json",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "memory-chart-invariants",
      "phase": "phase-parallel",
      "status": "completed",
      "title": "Encode chart-rendering invariants in memory",
      "path": "memory/project_playbook_pipeline.md",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "memory-system-contract",
      "phase": "phase-1",
      "status": "completed",
      "title": "Write System Contract memory (roles of 5 properties, non-negotiable rules)",
      "path": "memory/project_system_contract.md",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "todo-page-mvp",
      "phase": "phase-1",
      "status": "completed",
      "title": "Build todo-list webpage MVP (this page)",
      "path": "/z/public_html/todos/",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "journal-user-steps",
      "phase": "phase-1",
      "status": "pending",
      "title": "USER: 3 steps to finish journal persistence",
      "owner": "user",
      "priority": "P0",
      "note": "1) Run add_scorecard_columns.sql  2) Confirm user_id=2 and db=trsports_tradingjournal_dev match your server (now pre-set in save_journal.php from gen_scorecard.py)  3) Regenerate a scorecard, click Save, look for green Saved-to-DB",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "restore-0416",
      "phase": "phase-1",
      "status": "pending",
      "title": "Restore 4/16 rescued notes into DB",
      "owner": "claude",
      "note": "SQL insert after user finishes the 3 steps above.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "rescue-localstorage",
      "phase": "phase-1",
      "status": "pending",
      "title": "USER: rescue any remaining localStorage scorecard_journal_* keys",
      "owner": "user",
      "note": "F12 -> Application -> Local Storage -> trsportsacademy.com -> copy keys.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "diag-indicator-alerts",
      "phase": "phase-2",
      "status": "completed",
      "title": "Investigate indicator_alerts table: DDL, ingestion path, why empty",
      "owner": "claude",
      "priority": "P1",
      "note": "RESOLVED 4/18: table is alerts_tradingview.indicator_alerts (NOT main.*); it has data; gen_reports.py queries it fine. Scorecard 0/98 is a client-side symptom -- enrich_trades_from_db returns {} when no SSH tunnel to localhost:5432. Fix: run gen_scorecard.py ON a2, or open tunnel before running locally.",
      "completed_at": "2026-04-18",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "verify-enrich-on-server",
      "phase": "phase-2",
      "status": "pending",
      "title": "Re-run gen_scorecard.py ON a2 (or via tunnel) and confirm enrichment populates",
      "owner": "user",
      "priority": "P1",
      "note": "Expected: 'DB enrichment: N entries with regime/playbook tags' in log, non-zero enriched count on scorecard.",
      "created_at": "2026-04-18",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "audit-webhook-coverage",
      "phase": "phase-2",
      "status": "pending",
      "title": "Audit what alerts_tradingview.indicator_alerts is MISSING vs what the scorecard wants (stop, tp1, tp2, rr, grade, rvol, ema_slope)",
      "owner": "claude",
      "priority": "P1",
      "note": "Columns present: entry_price, vwap, vwap_upper_1..3, vwap_lower_1..3, band_spread_1..3, band_spread_rate, band_slope, market_regime, playbook_type, strategy, direction, indicator, exit_price, exit_reason. Likely missing: stop_price, tp1_price, tp2_price, pre_trade_grade, rvol, ema_slope. Check full DDL before assuming.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "wire-webhooks",
      "phase": "phase-2",
      "status": "pending",
      "title": "Add missing webhook fields (stop/tp/grade/rvol/ema_slope) to alerts_tradingview.indicator_alerts + TradingView alert JSON",
      "owner": "claude",
      "priority": "P1",
      "note": "Only needed for fields the audit above flags as missing. Do NOT rebuild the pipe -- it works.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "remove-na-labels",
      "phase": "phase-2",
      "status": "pending",
      "title": "Remove webhook-data N/A placeholders from scorecard once webhooks flow",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "smb-rubric-grading",
      "phase": "phase-2",
      "status": "pending",
      "title": "Per-trade grading: apply SMB Pre-Trade Rubric",
      "owner": "claude",
      "note": "Setup Context, Entry Trigger, Trade Management. Enables letter grade per trade.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "overtrading-detect",
      "phase": "phase-2",
      "status": "pending",
      "title": "Overtrading / chase detection",
      "owner": "claude",
      "note": "Flag clusters of fills above Band 1 like the IBIT 11:20-11:50 pattern.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "per-trade-charts",
      "phase": "phase-3",
      "status": "pending",
      "title": "Per-trade execution charts on scorecard",
      "owner": "claude",
      "note": "Thumbnail in row, lightbox expand. ET + session-only + stop/tp lines.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "multi-trader-scorecards",
      "phase": "phase-4",
      "status": "pending",
      "title": "Multi-trader scorecards (Enrique, Preston, Nolan, Brian, Thomas)",
      "owner": "claude",
      "note": "Parameterize gen_scorecard.py --user. Namespace output. Trader selector on index.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "shared-user-dropdown",
      "phase": "phase-2_5",
      "status": "pending",
      "title": "Shared user dropdown: query users-with-data, drive every page from ?user= URL param",
      "owner": "claude",
      "priority": "P1",
      "note": "One query joining main.users with main.journal + indicator_alerts, counts days_with_trades + alerts_received. Feeds all 5 property pages.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "project": "ams"
    },
    {
      "id": "add-user-role-approval",
      "phase": "phase-2_5",
      "status": "pending",
      "title": "Add role + is_approved columns to main.users",
      "owner": "claude",
      "priority": "P1",
      "note": "Enables admin/trader distinction and approve/deny gate. Migration: ALTER TABLE main.users ADD COLUMN role VARCHAR(20) DEFAULT 'trader', ADD COLUMN is_approved BOOLEAN DEFAULT false.",
      "created_at": "2026-04-18",
      "property": "foundation",
      "project": "ams"
    },
    {
      "id": "identity-resolution-fn",
      "phase": "phase-2_5",
      "status": "completed",
      "title": "Write identity resolution function: account_id -> user_id -> trader_name",
      "owner": "claude",
      "note": "CLOSED 2026-05-08 \u00e2\u20ac\u201d duplicate of p0-identity-fn (Foundation P0 task with same Python+PHP resolve_user helper). Merged for clarity. Original note retained: One shared helper in Python + PHP. Every generator and endpoint imports it. Replaces hardcoded user_id=2 in gen_scorecard.py.",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "completed_at": "2026-05-08",
      "superseded_by": "p0-identity-fn",
      "project": "ams"
    },
    {
      "id": "param-surface-strategy-registry",
      "phase": "phase-5",
      "status": "pending",
      "title": "Pivot Param Surface into canonical strategy registry (VWAP bands = row 1, user strategies = rows 2..N)",
      "owner": "claude",
      "priority": "P1",
      "note": "Every strategy gets: name, json_rules, owner, backtest_stats, live_performance. All trades get scored against every registered strategy.",
      "created_at": "2026-04-18",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "param-surface-freshness",
      "phase": "phase-5",
      "status": "pending",
      "title": "Param Surface data freshness check: every strategy view shows last_trade_ingested timestamp + alert if stale",
      "owner": "claude",
      "priority": "P1",
      "note": "Directly ties to data-freshness-audit. Every page in the registry shows an age badge; red if >24h stale on a trading day.",
      "created_at": "2026-04-18",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "strategy-library-ddl",
      "phase": "phase-5",
      "status": "pending",
      "title": "Create main.user_strategies table (id, user_id, name, json_rules, is_approved)",
      "owner": "claude",
      "note": "DDL migration. FK to main.users. This is the storage behind Param Surface rows 2..N.",
      "created_at": "2026-04-18",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "strategy-upload-ui",
      "phase": "phase-5",
      "status": "pending",
      "title": "Strategy upload UI: users submit JSON rules, admin approves",
      "owner": "claude",
      "note": "STATUS 2026-05-06: Static reference page /strategies/ shipped 4/30. Outstanding: submit form + admin approve queue at /strategies/admin/, gated by role=admin. Blocked on phase-2_5 (add-user-role-approval).",
      "created_at": "2026-04-18",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "strategy-backtest",
      "phase": "phase-5",
      "status": "pending",
      "title": "Backtest uploaded strategy against user's own trades",
      "owner": "claude",
      "note": "Reuses kti backtest-* engine. 'My strat on DB data' view.",
      "created_at": "2026-04-18",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "me-vs-firm",
      "phase": "phase-5",
      "status": "pending",
      "title": "'Me vs Firm' peer comparison toggle on Scorecards + Monthly",
      "owner": "claude",
      "note": "Adds peer-aggregate column: your win rate vs firm median, your expectancy vs firm median, etc.",
      "created_at": "2026-04-18",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "system-hub",
      "phase": "phase-1",
      "status": "completed",
      "title": "Architecture hub at /system/",
      "path": "/z/public_html/system/index.html",
      "completed_at": "2026-04-18",
      "note": "Single-page system overview with flow diagram, property list, data layer schema, rules, status dots. Paste into new Claude sessions.",
      "created_at": "2026-04-18",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "mgp-find-cmt-breadth",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Locate Ned Davis breadth model chapter in CMT materials",
      "owner": "claude",
      "priority": "P1",
      "note": "CLOSED 2026-05-08 \u00e2\u20ac\u201d superseded by phase-3-breadth-live (completed 2026-04-19) which computes breadth internally from KITE top-100 SP500 constituents (8-indicator composite). The Ned Davis chapter dependency was bypassed by going to source data; no longer need to locate it. Original note retained: Search Google Drive CMT folder for breadth / advance-decline / McClellan / Ned Davis chapter. Extract the breadth indicators + scoring rules.",
      "created_at": "2026-04-18",
      "property": "mgp",
      "completed_at": "2026-05-08",
      "superseded_by": "phase-3-breadth-live",
      "project": "atlas"
    },
    {
      "id": "mgp-template",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Draft MGP template: pre-market sections keyed off SMB emails + breadth score",
      "owner": "claude",
      "priority": "P1",
      "note": "SUPERSEDED by mgp-live. Superseded \u00e2\u20ac\u201d mgp_template.html shipped with 7 sections + calendar nav",
      "completed_at": "2026-04-21",
      "created_at": "2026-04-18",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-email-ingest",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "SMB email ingest: scan morning emails, extract symbols + setup notes",
      "owner": "claude",
      "note": "SUPERSEDED by mgp-live. Superseded \u00e2\u20ac\u201d vital_dawn_parser.py ingests SMB Vital Dawn emails directly",
      "completed_at": "2026-04-21",
      "created_at": "2026-04-18",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-breadth-feed",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Ned Davis breadth data feed",
      "owner": "user",
      "note": "SUPERSEDED by phase-3-breadth-live. Superseded \u00e2\u20ac\u201d breadth computed internally from KITE SP500 constituents, no external feed needed",
      "completed_at": "2026-04-21",
      "created_at": "2026-04-18",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-gen-script",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Write gen_mgp.py: template + email_json + breadth_score -> HTML",
      "owner": "claude",
      "note": "SUPERSEDED by mgp-live. Superseded \u00e2\u20ac\u201d gen_mgp.py shipped and runs 05:47 ET weekdays",
      "completed_at": "2026-04-21",
      "created_at": "2026-04-18",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-register-hub",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Register MGP as 7th property on system hub + add to top nav",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "Done \u00e2\u20ac\u201d tile live on /system/ as LIVE. Generator gen_mgp.py runs pre-market (05:47 ET weekdays).",
      "created_at": "2026-04-18",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-live",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "MGP live \u00e2\u20ac\u201d 7 sections rendered from Vital Dawn + KITE + Alpha Trend",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "Sections: Index Bias, Leadership & Rotation (Alpha Trend with RSI bars), Ned Davis breadth (placeholder), Stocks In Play (SMB template), Flagged Names, Earnings, Econ Calendar, Playbook Pick. Calendar prev/next nav + date dropdown. Daily at 05:47 ET via scheduled task mgp-daily.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-breadth-live",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Populate Ned Davis breadth (Section 3) with real data from SPX constituents via KITE",
      "owner": "claude",
      "priority": "P1",
      "note": "SUPERSEDED by phase-3-breadth-live. Superseded \u00e2\u20ac\u201d breadth live from KITE top-100 constituents; composite renders with real values; cache at breadth_cache/breadth_YYYY-MM-DD.json",
      "completed_at": "2026-04-21",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-multi-tf-bias",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Multi-timeframe ChipHazzard bias engine (SPY/QQQ/IWM \u00c3\u2014 1D/66m/15m/5m)",
      "owner": "claude",
      "note": "Section 1 enrichment. Compute ChipHazzard states from the Pine v2.2 indicator on 4 timeframes per index. Stacked alignment = high-confidence bias.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "mgp-bias-reconciliation",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Vital Dawn narrative vs our technicals \u00e2\u20ac\u201d track which was right",
      "owner": "claude",
      "path": "/z/setup_daily_reporting/reports/convergence.py",
      "completed_at": "2026-04-26",
      "note": "Shipped \u00e2\u20ac\u201d convergence.py module compares ChipHazzard alignment + breadth composite (our bias) vs Vital Dawn futures-mean bias (their bias) into 3 buckets, emits banner on MGP. Today's MGP shows partial-alignment banner: tech 'lean-risk-on' vs VD 'unknown'. Weekly win/loss reconciliation tile still TBD.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-daily-stop-pipe",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Daily Stop: wire from Accountability page into MGP Max Risk calc",
      "owner": "claude",
      "priority": "P1",
      "note": "STATUS 2026-05-06: Daily Stop standalone tool verified live at /daily-stop/. API endpoint phase-2-daily-stop-live shipped at /scorecards/api/user_settings.php. Outstanding: wire MGP Max Risk to read daily_stop from main.user_settings via that endpoint instead of hardcoded $500.",
      "created_at": "2026-04-19",
      "property": "daily_stop",
      "project": "acct_v2"
    },
    {
      "id": "mgp-sip-db",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Persist Stocks In Play to DB for scorecard join",
      "owner": "claude",
      "note": "New table alerts_tradingview.stocks_in_play_daily. Keyed by (trade_date, ticker). Scorecard joins trades to it on (trade_date, symbol) \u00e2\u2020\u2019 Planned vs Actual.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "multi_trader"
    },
    {
      "id": "mgp-rsi-overextended-rule",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Overextended bucket: adopt James's playbook rule (RSI>70 AND RSI-MA10 \u00e2\u02c6\u02c6 60-70)",
      "owner": "claude",
      "note": "Currently uses simple RSI \u00e2\u2030\u00a5 75. James's rule distinguishes spike-into-overbought (mean-reversion risk) from persistent strong trend. Add rsi_ma10 to the computation + split Overextended into 'Spike Extended' vs 'Persistent Trend' buckets.",
      "created_at": "2026-04-19",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "mgp-strategy-templates",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "IF/THEN templates align to named strategies (VWAP Pull-In, Back$ide Scalp, Fash Late variants)",
      "owner": "claude",
      "priority": "P1",
      "note": "Currently generic short/long templates. Replace with one template per named setup. VWAP Pull-In/Back$ide Scalp = MGP-level pattern classifications inferred from webhook indicator + direction + entry_type + gap context (user confirmed option B). Fash Late has 13+ SMB_FASH_EVT_* sub-variants to map.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-conservative-moderate-aggressive",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Rename Stocks In Play grades A/B/C \u00e2\u2020\u2019 Conservative/Moderate/Aggressive",
      "owner": "claude",
      "note": "Matches alerts_tradingview.indicator_alerts.entry_type already in use. Same risk allocation mapping (Aggressive=30%, Moderate=20%, Conservative=10% of daily stop).",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "multi_trader"
    },
    {
      "id": "mgp-key-level-enrichment",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Enrich Stocks In Play key levels with full webhook-carried set",
      "owner": "claude",
      "note": "indicator_alerts already has: VWAP bands 1/2/3 upper+lower, DMA 2d/5d/10d/20d/50d/200d, EMA(9), pivot_points jsonb, ATR. Use all instead of just PDH/PDL/PDC.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "multi_trader"
    },
    {
      "id": "strategy-vpi-rules",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "VWAP Pull-In: codify rules (Claude + KITE, no webhook)",
      "owner": "claude",
      "priority": "P1",
      "note": "User confirmed VWAP Pull-In is a Claude/KITE-built classification, not a webhook. Needs: trigger rules (gap + MA stack + VWAP state), entry grades (Conservative/Moderate/Aggressive), exit logic. Reference: compare/contrast vs FL2 BandPullback at /backtests/fl2-band-trail/.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "strat_lib"
    },
    {
      "id": "strategy-vpi-backtest",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "VWAP Pull-In: walk-forward backtest",
      "owner": "claude",
      "priority": "P1",
      "note": "Once rules are codified. Same walk-forward structure as fl2-band-trail. Publish to /backtests/vwap-pull-in/.",
      "created_at": "2026-04-19",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "strategy-backside-rules",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Back$ide Scalp: codify rules (Claude + KITE, no webhook)",
      "owner": "claude",
      "note": "Gap-up failure pattern: sell-off into PDH, 2-day VWAP, or DMA level. IF pullback to prior day VWAP / 2-day VWAP, monitor for backside into VWAP. Then next trade = Fash Late style re-entry (kept separate).",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "strat_lib"
    },
    {
      "id": "strategy-backside-backtest",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Back$ide Scalp: walk-forward backtest",
      "owner": "claude",
      "created_at": "2026-04-19",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "strategy-align-scorecard-pie",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Align Scorecard Strategy Breakdown pie labels to the new strategy names",
      "owner": "claude",
      "note": "Pie currently shows: Chase/Extended, VWAP Pull-In, Back$ide Scalp, VWAP Pull-In \u00e2\u2020\u2019 Trend. 'Chase/Extended' should be a warning class, not a strategy. Align to codified strategies.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "strat_lib"
    },
    {
      "id": "strategy-planned-vs-actual",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Scorecard: 'Planned vs Actual' tile \u00e2\u20ac\u201d did trades match the MGP Stocks In Play plan?",
      "owner": "claude",
      "note": "Join trades \u00e2\u2020\u2019 stocks_in_play_daily on (trade_date, symbol). Per pick show: did you trade it? entry/stop/target adherence? Did A-grade picks win more than C-grade?",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "strategy-daily-stop-used",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Scorecard: 'Daily Stop % Used' gauge",
      "owner": "claude",
      "note": "abs(day_loss) / daily_stop \u00c3\u2014 100. Red if > 100%. Ties scorecard to accountability rule.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "strategy-followed-plan",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "title": "Per-trade: 'Did you follow your plan?' flag on scorecard row",
      "owner": "claude",
      "note": "Given the MGP IF/THEN for this ticker, did the actual entry/stop/target match? boolean + delta.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "pm-nightly-task",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Nightly 11PM project-manager: scan Z drive + write daily writeup",
      "owner": "claude",
      "priority": "P1",
      "note": "STATUS 2026-05-06: PARTIAL \u00e2\u20ac\u201d nightly task firing and producing /todos/daily/YYYY-MM-DD.md writeups (verified through 2026-05-06). Z drive scan + Claude-session summary + writeup ALL working. Two pieces still missing: (4) auto-update todos.json task statuses on detected completions, (5) visual progress calendar (tracked separately as pm-calendar-view).",
      "created_at": "2026-04-19",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "pm-calendar-view",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Progress calendar view on /todos/",
      "owner": "claude",
      "note": "Heatmap calendar \u00e2\u20ac\u201d each cell = # of tasks completed that day, tooltip shows summary. Like GitHub contributions. Click day \u00e2\u2020\u2019 daily writeup.",
      "created_at": "2026-04-19",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "pm-priority-viz",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Visual: what was the priority + where did the work land?",
      "owner": "claude",
      "note": "Per-day breakdown pie/bar: % time on MGP vs scorecard vs strategy vs breadth, with which files were touched.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "ux-cross-nav",
      "phase": "phase-0",
      "status": "pending",
      "title": "Cross-page navigation: every property must link to every other property",
      "owner": "claude",
      "priority": "P1",
      "note": "Currently clicking into scorecards/playbooks/backtests/monthly has no way back to the others \u00e2\u20ac\u201d only /mgp/ has the full top-nav. Unify the top-nav component across ALL properties (System Hub, Todos, Scorecards, Playbooks, Monthly, Param Surface, Accountability, Backtests). Use a single included HTML partial or a PHP include.",
      "created_at": "2026-04-19",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "scorecard-time-of-day-equity",
      "phase": "phase-3",
      "status": "pending",
      "title": "Scorecard Intraday Equity Curve: x-axis = time of day (not Trade #)",
      "owner": "claude",
      "priority": "P1",
      "note": "Currently shows Trade 1..101 on x-axis which loses the when-did-I-make-money signal. Switch to time-of-day (09:30 \u00e2\u2020\u2019 16:00 ET). Same y-axis (cumulative P&L). Needs entry_time from trade dict; if missing from some trades interpolate or show gap.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "scorecard-strategy-time-of-day",
      "phase": "phase-3",
      "status": "pending",
      "title": "Strategy Breakdown: add time-of-day dimension + per-strategy correlation",
      "owner": "claude",
      "priority": "P1",
      "note": "Right now pie shows total share per strategy. Extend: (1) stacked bar by 30-min bucket (09:30, 10:00, ...) showing which strategy is active when; (2) correlation matrix (Chase/Extended vs VWAP Pull-In vs Back$ide Scalp vs VWAP Pull-In->Trend) \u00e2\u20ac\u201d Pearson correlation of per-day P&L across the 4 strategies over trailing 30/60 days. Highlights which strategies tend to win/lose together.",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "mgp-bias-multi-tf-risk",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Consider splitting Overall Bias into Daily / Intermediate / Short-term TF buckets",
      "owner": "claude",
      "note": "User asked 2026-04-19 whether to split into daily vs intermediate vs small-TF risk. Trade-off: more info vs more visual clutter. Prototype with collapsible mini-sections inside the bias card \u00e2\u20ac\u201d one header showing composite, expandable to per-TF state.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-bias-wtd-daily",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Overall Bias card: add WTD gain and Daily gain stats",
      "owner": "claude",
      "note": "User asked to add WTD and Daily on the bias card. Alongside futures bps \u00e2\u20ac\u201d shows where we are NOW vs where we were Monday open. Start simple: just SPY/QQQ/IWM WTD% and day change%.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-breadth-click-indicator",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Breadth indicators: click tile to reveal historical chart of that indicator",
      "owner": "claude",
      "note": "Each of the 8 breadth cells becomes clickable \u00e2\u20ac\u201d opens a mini chart showing the indicator's trend over last 60/252 days + threshold zones. Lightweight-charts. First time user has historical context, not just today's snapshot.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-vital-dawn-ticker-charts",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Vital Dawn Flagged Names: auto-pull chart + generate IF/THEN for each",
      "owner": "claude",
      "note": "Currently flagged names are just a table with text context. Should pull daily + intraday chart per ticker and generate IF/THEN plan using same rule engine as Stocks In Play. Basically: promote every flagged ticker into a full Stocks In Play card, or at minimum chart + 1-line levels + 1-line IF/THEN.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-multi-tf-chart-viewer",
      "phase": "phase-mgp",
      "status": "pending",
      "title": "Multi-timeframe chart viewer (click ticker \u00e2\u2020\u2019 modal with 1D/66m/15m/5m charts)",
      "owner": "claude",
      "note": "Site-wide capability. Click any ticker on MGP/Scorecards/Playbooks \u00e2\u2020\u2019 opens modal with 4 stacked charts at different TFs. Same KITE data, same bar-fetch helper. Makes post-automation review trivial.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "phase-5a-sip-db",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Phase 5a: stocks_in_play_daily DB table + persist from gen_mgp.py",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "Done. Scorecard joins trades -> morning picks on (trade_date, symbol).",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "phase-5b-enrichments",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Phase 5b: scorecard_enrichments.py module (daily stop gauge, planned vs actual, follow-plan badge)",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "created_at": "2026-04-19",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "strategy-vpi-live",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "title": "VWAP Pull-In walk-forward backtest published at /backtests/vwap-pull-in/",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "80 VPI-zone entries analyzed from indicator_alerts over 30 days. Spearman 0.28 (moderate stability). Best IS combo LB3_UB3 ranks #3 OOS. Deflated Sharpe 0 = more data needed. Reuses gen_walkforward.py framework with +/-0.5 SD filter from regime_classifier.",
      "created_at": "2026-04-19",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "phase-3-breadth-live",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Ned Davis breadth composite live from KITE top-100 constituents",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "8 indicators computed daily. 4/17 snapshot: composite +5 (risk-on). Cached to breadth_cache/breadth_<date>.json. Universe in breadth_universe.py (~120 large caps, ~80% S&P weight).",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "phase-4-chiphazzard-v1",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "ChipHazzard index stack panel (SPY/QQQ/IWM, 1D)",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "Port of Pine v2.2 core state logic (markup/anticipate_up/anticipate_decline/exit_long/avoid). Renders on MGP Section 1. Multi-TF (66m/15m/5m) tracked as separate todo for v2.",
      "created_at": "2026-04-19",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "phase-2-daily-stop-live",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Daily Stop pipe: DB-backed main.user_settings, shared between Accountability and MGP",
      "owner": "claude",
      "completed_at": "2026-04-19",
      "note": "Endpoint: /scorecards/api/user_settings.php (GET/POST, CORS-enabled for smb subdomain). MGP Stocks In Play Max Risk now reflects $130 daily stop. Accountability page needs separate work to wire save/load to this endpoint.",
      "created_at": "2026-04-19",
      "property": "daily_stop",
      "project": "acct_v2"
    },
    {
      "id": "fix-stacked-axis",
      "phase": "phase-parallel",
      "status": "pending",
      "title": "Playbook chart: stacked top/bottom x-axes must share visible time range",
      "owner": "claude",
      "note": "exec<->cumPnL, daily<->RSI panels.",
      "created_at": "2026-04-18",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "fix-session-only",
      "phase": "phase-parallel",
      "status": "pending",
      "title": "Playbook chart: intraday must render session-only, no gaps on 15-min",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "enforce-et-tz",
      "phase": "phase-parallel",
      "status": "pending",
      "title": "Enforce America/New_York timezone across all chart + data timestamps",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "playbooks",
      "project": "atlas"
    },
    {
      "id": "chart-qa-guardrail",
      "phase": "phase-parallel",
      "status": "pending",
      "title": "Add pre-publish chart QA guardrail to generators",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "playbooks",
      "project": "atlas"
    },
    {
      "id": "data-freshness-audit",
      "phase": "phase-1",
      "status": "pending",
      "title": "DATA FRESHNESS: audit why newest trades aren't flowing into DB + Param Surface",
      "owner": "claude",
      "priority": "P0",
      "note": "User reports newest trades not updating. Check: (1) last trade_date in main.journal vs today, (2) last datetime_ny in alerts_tradingview.indicator_alerts, (3) Gr8Trade ingestion cron status on a2, (4) pipeline.js last run log. This is THE spine breathing -- blocks Scorecards, Monthly, Param Surface, MGP.",
      "created_at": "2026-04-19",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "fix-pipeline-0417",
      "phase": "phase-1",
      "status": "pending",
      "title": "Fix pipeline not running for 2026-04-17 (subset of freshness audit)",
      "owner": "claude",
      "priority": "P0",
      "created_at": "2026-04-19",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "fix-missing-days",
      "phase": "phase-1",
      "status": "pending",
      "title": "Fix scorecard manifest missing 4/1-4/13 days (likely same root cause)",
      "owner": "claude",
      "priority": "P0",
      "created_at": "2026-04-19",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "retire-trade-executions",
      "phase": "phase-hold",
      "status": "cancelled",
      "title": "Retire /db_reports/trade_executions/ dead page",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "foundation",
      "cancelled_at": "2026-05-08",
      "cancelled_reason": "Low-priority cleanup of /db_reports/trade_executions/ dead page. Not worth active cycles. Page is harmless if left alone \u00e2\u20ac\u201d no inbound links from current properties. Reopen if a future audit demands a deletion.",
      "note": "CANCELLED 2026-05-08 \u00e2\u20ac\u201d Low-priority cleanup of /db_reports/trade_executions/ dead page. Not worth active cycles. Page is harmless if left alone \u00e2\u20ac\u201d no inbound links from current properties. Reopen if a future audit demands a deletion.",
      "project": "atlas"
    },
    {
      "id": "confirm-monthly",
      "phase": "phase-hold",
      "status": "completed",
      "title": "Confirm April monthly report auto-upload is correct",
      "owner": "user",
      "created_at": "2026-04-18",
      "property": "monthly_reports",
      "completed_at": "2026-05-08",
      "note": "CLOSED 2026-05-08 \u00e2\u20ac\u201d April 2026 monthly report has been auto-regenerating daily via run_all_eod.bat since 4/30; April-2026 page confirmed live at https://smb.a1prospect.com/db_reports/james-monthly-reports/april-2026/. The original ad-hoc confirmation question is moot. Original note retained: ",
      "project": "atlas"
    },
    {
      "id": "cross-link",
      "phase": "phase-hold",
      "status": "cancelled",
      "title": "Cross-link scorecard / monthly / strategy analysis",
      "owner": "claude",
      "created_at": "2026-04-18",
      "property": "scorecards",
      "cancelled_at": "2026-05-08",
      "cancelled_reason": "Vague task ('Cross-link scorecard / monthly / strategy analysis') \u00e2\u20ac\u201d no concrete spec, no defined deliverable. Cross-property navigation is now handled by the system hub + the per-property top-nav (nav.js), so the implicit goal is satisfied through that path. Reopen with a specific link / page that's missing if/when one is identified.",
      "note": "CANCELLED 2026-05-08 \u00e2\u20ac\u201d Vague task ('Cross-link scorecard / monthly / strategy analysis') \u00e2\u20ac\u201d no concrete spec, no defined deliverable. Cross-property navigation is now handled by the system hub + the per-property top-nav (nav.js), so the implicit goal is satisfied through that path. Reopen with a specific link / page that's missing if/when one is identified.",
      "project": "multi_trader"
    },
    {
      "id": "mighty-rider",
      "phase": "phase-hold",
      "status": "cancelled",
      "title": "Investigate MIGHTY RIDE PLAYER injection if it reappears",
      "owner": "user",
      "created_at": "2026-04-18",
      "property": "foundation",
      "cancelled_at": "2026-05-08",
      "cancelled_reason": "Watch-only item, not a tracked project. Has not reappeared in 20+ days. Moved to memory file watchlist_mighty_rider.md with detection + response steps if it ever returns.",
      "note": "CANCELLED 2026-05-08 \u00e2\u20ac\u201d Watch-only item, not a tracked project. Has not reappeared in 20+ days. Moved to memory file watchlist_mighty_rider.md with detection + response steps if it ever returns.",
      "project": "atlas"
    },
    {
      "id": "todos-snapshots",
      "phase": "phase-hold",
      "status": "completed",
      "title": "Todo snapshots + morning/evening reports",
      "owner": "claude",
      "note": "CLOSED 2026-05-08 \u00e2\u20ac\u201d shipped as the `todos-scan-twice-daily` scheduled-tasks routine (cron: 0 5,17 * * *), enabled, lastRunAt confirmed in the scheduled-tasks system. Output to \\\\a1prospect.com@SSL@2078\\DavWWWRoot\\james\\todos_snapshot.md. Original note retained: Next increment after todo page MVP.",
      "created_at": "2026-04-19",
      "property": "pm_system",
      "completed_at": "2026-05-08",
      "superseded_by": "scheduled-task:todos-scan-twice-daily",
      "project": "atlas"
    },
    {
      "id": "mgp-multi-tf-risk-lanes",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Multi-TF Risk Read: 15m/30m/65m intraday lanes populated",
      "path": "/z/setup_daily_reporting/reports/chiphazzard.py",
      "completed_at": "2026-04-21",
      "note": "Added classify_index_panel_intraday() + 5m -> 15m/30m/65m aggregator. MGP Multi-TF Risk Read now shows all 4 lanes with per-index state chips.",
      "created_at": "2026-04-20",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "sc-missed-opportunity-sim",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Planned vs Actual: Would-Have column on untraded picks + Missed Winners tile",
      "path": "/z/setup_daily_reporting/reports/scorecard_enrichments.py",
      "completed_at": "2026-04-21",
      "note": "simulate_untraded_pick() walks 5-min bars for each untraded pick; races target vs stop from RTH open. Planned vs Actual adds Would-Have column + Missed Winners / Correctly Skipped tiles. Separates plan edge from execution.",
      "created_at": "2026-04-20",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "pm-activity-scanner-deploy",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Deploy z_activity_scanner.py on each Windows machine via Task Scheduler",
      "owner": "user",
      "priority": "P1",
      "created_at": "2026-04-21",
      "note": "Script at Z:\\public_html\u0007ctivity\\z_activity_scanner.py. Import task_scheduler_import.xml on each PC. Runs every 15min, writes COMPUTERNAME-HHMM.json to /activity/YYYY-MM-DD/. User must deploy on DESKTOP-AKU8I62 + any other machines.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "pm-activity-aggregate",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Nightly PM: aggregate per-computer activity logs into daily writeup",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-21",
      "note": "Update SKILL.md to read /activity/YYYY-MM-DD/*.json, group files by computer, annotate daily report with which machine did what. Enables multi-computer attribution.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "pm-computers-dashboard",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "title": "Active computers panel: /activity/heartbeat/*.json \u00e2\u2020\u2019 freshness dots on system hub",
      "owner": "claude",
      "created_at": "2026-04-21",
      "note": "Each machine writes heartbeat every 15min. System hub reads them and shows per-computer last-seen + area-active. Stale if > 20min.",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "risk-hud-stop-gauge",
      "phase": "phase-0",
      "status": "pending",
      "title": "Accountability HUD: stop-utilization gauge bars on each trader card",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-21",
      "note": "Visual fuel gauge: abs(blotter) / daily_stop * 100%. Color: green <50%, yellow 50-80%, orange 80-99%, red >=100% (breach). Show % used + $ remaining before stop. Replaces the thin color bar at top of each card.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "risk-hud-summary-row",
      "phase": "phase-0",
      "status": "pending",
      "title": "Accountability HUD: risk summary row \u00e2\u20ac\u201d # breached, # danger, # warning, # clear",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-21",
      "note": "Sticky header tile row: 4 counters (Clear / Warning 50-80% / Danger 80-99% / Breach). Risk manager can scan status in one glance without reading individual cards. Clicking a tile filters the cards below.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "risk-hud-sort-risk",
      "phase": "phase-0",
      "status": "pending",
      "title": "Accountability HUD: Risk Manager view \u00e2\u20ac\u201d sort cards by stop% (worst first) + risk rank badge",
      "owner": "claude",
      "created_at": "2026-04-21",
      "note": "Toggle \"Risk View\" button. Resorts all cards by descending stop utilization. Adds rank badge (#1 most at risk). Most urgent traders always top-left. Normal view preserves alphabetical.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "risk-hud-auto-refresh",
      "phase": "phase-0",
      "status": "pending",
      "title": "Accountability HUD: auto-refresh every 60s in Risk Manager view",
      "owner": "claude",
      "created_at": "2026-04-21",
      "note": "Risk manager leaves the tab open during the trading day. Page re-fetches data every 60 seconds. Countdown timer visible. Flashing pulse on any card that changes status tier since last refresh.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "risk-hud-api-wire",
      "phase": "phase-0",
      "status": "pending",
      "title": "Wire Risk Manager page to smb.a1prospect.com live API (replace mock data)",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-21",
      "path": "/z/public_html/accountability/risk.html",
      "note": "risk.html built with full gauge UI. Set API_URL at top to the smb subdomain endpoint that returns { traders: [{name, blotter, daily_stop, trades, win_rate, mult_mtm, violations_week}] }. Then remove MOCK_DATA.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "pm-bootstrap-prompt",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "title": "Write PC bootstrap prompt for any machine to join the PM system",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-21",
      "completed_at": "2026-04-21",
      "path": "/z/public_html/activity/PC_BOOTSTRAP_PROMPT.md",
      "note": "Copy prompt into any Claude session. Sets up activity scanner, briefs Claude on project context, defines morning/afternoon/EOD commands.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "run-scientific-layer-migrations",
      "phase": "phase-0",
      "status": "pending",
      "title": "USER: run scientific_layer migrations 001\u00e2\u20ac\u201c009 via psql tunnel",
      "owner": "user",
      "priority": "P0",
      "created_at": "2026-04-22",
      "path": "/z/setup_daily_reporting/reports/migrations/scientific_layer/README.md",
      "note": "Run in order: set_db.sh start, then psql -f 001...009 against trsports_tradingjournal_dev. Unlocks: p0-daily-summary-view (3/6), sizing calculator DB wiring, rolling_stats, setup_performance, sizing_recommendation views. All migrations are additive \u00e2\u20ac\u201d no drops, no data changes. Rollback: rollback_all.sql.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "scientific-layer-ui-shipped",
      "phase": "phase-5",
      "status": "completed",
      "title": "Scientific layer UI: Sizing Calculator + Strategy Bar v1.1.0 + playbook template stats",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-22",
      "completed_at": "2026-04-22",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/sizing/index.html",
      "note": "Sizing Calculator page (spec \u00c2\u00a75): grade \u00c3\u2014 edge-overlay \u00c3\u2014 DD-taper \u00e2\u2020\u2019 live share count. Strategy Bar bumped v1.1.0 with computeSizing/gradeMult/edgeOverlay/ddTaper. Playbook template gets 4-up Scientific-Layer stat row + tooltip labels. VWAP Bands hub gets Scientific Layer card + nav links. DB wiring pending migrations.",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "scientific-overlay-per-trade",
      "phase": "phase-5",
      "status": "completed",
      "title": "Per-trade edge overlay on every scorecard (16-strategy catalog with verdict tags)",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/public_html/scorecards/_scientific_overlay.js",
      "note": "Self-contained drop-in JS injected into 14 historical scorecards. Walks every .playbook-card, looks up documented edge from inline catalog (16 setups: 7 VWAP family + 2 counter-trend + 3 breakout + 2 event-driven + 1 gap fade + Bella Fade alias). Renders Expectancy(R) / WR / MAE / Leave-on-table + verdict badge (PROVEN/WORKING/MARGINAL/NEGATIVE) with sizing-engine multiplier. Fixes /playbooks/ 404 with new landing page. Catalog will swap for /api/setup_performance once migration 006 lands.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "scientific-overlay-def-block",
      "phase": "phase-5",
      "status": "completed",
      "title": "\"What it is\" definition cards on scorecards + Bella Fade \u00e2\u2020\u201d Backside Scalp unification",
      "owner": "claude",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/public_html/scorecards/_scientific_overlay.js",
      "note": "Catalog extended with thesis/setup/entry/stop/target/bestWhen per strategy. ALIASES map resolves \"Bella Fade\", \"GGG\", \"Trader Dale\" \u00e2\u2020\u2019 canonical slugs. buildDefBlock() renders 4-row recipe card above edge-stats block. Backside Scalp shown as canonical with \"(also called Bella Fade)\" \u00e2\u20ac\u201d no data duplication.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "monthly-dashboard-v2",
      "phase": "phase-5",
      "status": "completed",
      "title": "Monthly Dashboard v2 (Scientific Layer attribution surface) \u00e2\u20ac\u201d root /james-monthly-reports/",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/setup_daily_reporting/reports/gen_monthly_dashboard.py",
      "note": "Spec \u00c2\u00a74.3 long-horizon attribution. Root /james-monthly-reports/ index was stale by 26 trading days (last point 3/31). New ~410-line generator pulls April-onward via kti trades range, aggregates daily P&L / by-symbol / WR / Sortino-60 / Calmar-60 / MDD. Replaced root index.html with: KPI strip \u00c2\u00b7 live equity curve with month markers \u00c2\u00b7 risk-adjusted ratios \u00c2\u00b7 attribution-by-symbol top-15 \u00c2\u00b7 best/worst trade cards \u00c2\u00b7 placeholders that light up on migrations 002/006 \u00c2\u00b7 drift detector. Wired into run_all_eod.bat as step 3/5. Behavioral-progression + setup-edge tables wait on migrations.",
      "property": "monthly_reports",
      "project": "atlas"
    },
    {
      "id": "mgp-pipeline-backup",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "MGP pipeline backup utility (timestamped ZIP of all generators + configs)",
      "owner": "claude",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/setup_daily_reporting/reports/backup_pipeline.py",
      "note": "Self-contained restore \u00e2\u20ac\u201d packages every Python module, template, config, and the server-side mailer.php. Excludes fixtures/outputs/secrets but bundles a sanitized template so the backup is rebuildable. Default writes to Z:\\setup_daily_reporting\\backups\\.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "mgp-mailer-infra",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Generic server-side mailer + MGP notify refresh (mailer.php + notify.py + recipients)",
      "owner": "claude",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/public_html/_api/mailer.php",
      "note": "Reusable POST endpoint at /_api/mailer.php \u00e2\u20ac\u201d any internal script can send via shared secret (uses local postfix, no SMTP creds). notify.py refactored into thin wrapper exposing send_email() generically; mgp_email.py + mgp_recipients.json refreshed to use it.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "shared-tooltip-dictionary",
      "phase": "phase-0",
      "status": "completed",
      "title": "Universal tooltip engine extended with market-index + breadth dictionary entries",
      "owner": "claude",
      "created_at": "2026-04-26",
      "completed_at": "2026-04-26",
      "path": "/z/public_html/_shared/tooltip.js",
      "note": "data-tt=\"key\" hover/tap tooltips for SPY, QQQ, IWM, A/D Line, Markup, Convergence (more entries to follow). Cross-property: any page including /_shared/tooltip.js gets the same definitions. Used by today's MGP convergence banner.",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "mgp-intraday-pipeline",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "Three-edition MGP pipeline: Morning + Mid-Day + Recap intraday generator",
      "owner": "claude",
      "priority": "P1",
      "created_at": "2026-04-27",
      "completed_at": "2026-04-27",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_intraday.py",
      "note": "Implements project_mgp_pipeline.md: Mid-Day (11:30 ET cutoff) + Recap (16:00 ET full-RTH) editions. Pulls Vital Mid-Day / Market Recap email, loads morning's persisted SIP picks from DB, walks 5-min bars to cutoff, classifies each pick (target/stop/open/no-data), identifies new tickers vs morning, emits per-edition HTML + email. Today's 2026-04-27-midday.html + recap.html rendered 19:17 ET.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-dynamic-calendar",
      "phase": "phase-mgp",
      "status": "completed",
      "title": "MGP dynamic calendar (mgp_cal.js) handles 3 editions per day from manifest",
      "owner": "claude",
      "created_at": "2026-04-27",
      "completed_at": "2026-04-27",
      "path": "/z/public_html/_shared/mgp_cal.js",
      "note": "Reads /mgp/manifest.json (now per-edition rows: morning|midday|recap), parses both URL forms, rebuilds Prev/Cur/Next/Jump/Latest on dated pages + latest banner + archive on /mgp/index.html. nav.js extended to auto-load it on /mgp/* pages. Old reports auto-learn about new ones.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "bella-fade-examples-generator",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Bella Fade worked-examples page generator (SNOW + HOOD reference)",
      "path": "/z/setup_daily_reporting/reports/bella_fade_examples/build_examples.py",
      "completed_at": "2026-04-28",
      "created_at": "2026-04-28",
      "note": "Becomes the model for Top 3 Wins/Losses on 5/1. Pulls SNOW + HOOD daily/intraday JSON, builds the canonical Bella Fade reference page at /playbooks/bella-fade-intern-template/examples.html.",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "mtf-vp-first-live-scans",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "MTF VP Scanner \u00e2\u20ac\u201d first live scans run end-to-end",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/mtf-vp/scans/",
      "completed_at": "2026-04-28",
      "created_at": "2026-04-28",
      "note": "Three completed scans land via Path A: QQQ (80.179 MODERATE), AAPL (8.749 BELOW_THRESHOLD), NVDA (96.153 ALIGNED_DOWN). First end-to-end on-demand scan run. Live TradingView ticket-in handler still pending.",
      "property": "mtf_vp_scanner",
      "project": "mtfvp"
    },
    {
      "id": "auto-classifier-backside-long",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Auto-classifier: BACKSIDE_LONG fires from live indicator_alerts (3 bugs killed)",
      "path": "/z/setup_daily_reporting/reports/regime_classifier.py",
      "completed_at": "2026-05-01",
      "created_at": "2026-05-01",
      "note": "regime_classifier.py + gen_scorecard.py + DB enrichment SQL all patched. Priority cascade BACKSIDE_LONG > BACKSIDE_SCALP > VPI->TREND > VPI > CHASE > UNCLASSIFIED. 60-min lookback CTE wired so rule-based path sees prior-extension context. 5/5 synthetic tests pass. Documented in SCIENTIFIC_SPEC_CHANGELOG.md addenda b/c.",
      "property": "scorecards",
      "project": "classifier"
    },
    {
      "id": "scorecard-dominant-tag-rendering",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "Scorecard symbol-row: dominant-tag rendering kills the IBIT 3-tag stack",
      "path": "/z/setup_daily_reporting/reports/gen_scorecard.py",
      "completed_at": "2026-05-01",
      "created_at": "2026-05-01",
      "note": "Counter() over per-symbol tags -> dominant tag only. Secondary tags with >=2 trades show as +N variant hint with hover tooltip. Single outliers roll into dominant tag. Documented as SCIENTIFIC_SPEC_CHANGELOG.md addendum d.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "best-april-2026-examples-page",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "Top 3 Wins + Top 3 Losses \u00e2\u20ac\u201d April 2026 examples page",
      "path": "/z/public_html/playbooks/best-april-2026/examples.html",
      "completed_at": "2026-05-01",
      "created_at": "2026-05-01",
      "note": "RTH-only re-aggregation per (Symbol, Day, Direction). Each card shows daily context, 1-min execution chart with all entry/exit fills as markers, strategy-bar against 5 playbooks, MGP daily-context tag. Drivers: best_trades_rth.py + build_best_april_examples.py. 7 tickers featured.",
      "property": "playbooks",
      "project": "atlas"
    },
    {
      "id": "me-vs-strategy-charts",
      "phase": "phase-3",
      "status": "completed",
      "owner": "claude",
      "title": "Me-vs-Strategy 2-minute Lightweight Charts on scorecard (3-view toggle)",
      "path": "/z/setup_daily_reporting/reports/gen_me_vs_strategy_charts.py",
      "completed_at": "2026-04-30",
      "created_at": "2026-04-30",
      "note": "Per-trade execution chart with Price+MAs / Trade Levels / VWAP+Bands toggle. Pulls 1m via kti, aggregates to 2m, computes session-anchored VWAP + 0.5/1.0/2.0 SD bands + 9/20/50/200 EMAs. Idempotent injector at inject_mvs_charts.py. Partial fulfillment of phase-3 per-trade-charts.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "strategy-library-page",
      "phase": "phase-5",
      "status": "completed",
      "owner": "claude",
      "title": "Strategy Library page \u00e2\u20ac\u201d canonical reference for all 16 strategies",
      "path": "/z/public_html/strategies/index.html",
      "completed_at": "2026-04-30",
      "created_at": "2026-04-30",
      "note": "/strategies/ landing page: filter bar, 5 family groups, per-card SVG band diagrams, 5-stat edge footers, verdict bars, deep-link #strat-<slug>. Static-only \u00e2\u20ac\u201d user upload + admin approve still pending (see strategy-upload-ui).",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "live-classifier-engine",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Live strategy-match HUD: per-symbol classify/<SYMBOL>.json + james-live.html",
      "path": "/z/public_html/scorecards/james-live.html",
      "completed_at": "2026-05-05",
      "created_at": "2026-05-05",
      "note": "14 per-symbol JSONs score live SD position against six playbooks. Each playbook returns TRIGGERED / TARGET_MET / NOT_APPLICABLE with gates_passed/total + canonical entry/stop/target + r_to_target_now. james-live.html is the HUD over the JSONs. Classifier finally has a live runtime.",
      "property": "live_classifier",
      "project": "classifier"
    },
    {
      "id": "mtf-trend-matrix",
      "phase": "phase-3",
      "status": "completed",
      "owner": "claude",
      "title": "Multi-Timeframe Trend Matrix \u00e2\u20ac\u201d 6-TF \u00c3\u2014 4-tier Points/Slope/Period score on scorecard symbol cards",
      "path": "/z/setup_daily_reporting/reports/mtf_trend.py",
      "completed_at": "2026-05-06",
      "created_at": "2026-05-06",
      "note": "Week/Day/65/30/15/5m \u00c3\u2014 Short/Int/Long/Conf score injected on every scorecard symbol card. Bressert Fig 10.4 + Shannon Table B-7. Live on scorecards 2026-05-06.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "daily-stop-tool",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Daily Stop tool \u00e2\u20ac\u201d /daily-stop/ trader control dashboard (max loss, trade caps, sizing, danger zone)",
      "path": "/z/public_html/daily-stop/",
      "completed_at": "2026-04-22",
      "created_at": "2026-04-22",
      "note": "Verified live 2026-05-06. Sets daily limits, tracks real-time P&L + win rate, logs trades, Danger Zone halts trading at max loss. Outstanding: register on /system/ hub and wire MGP Max Risk to read its setting.",
      "property": "daily_stop",
      "project": "acct_v2"
    },
    {
      "id": "scope-classifier-hud",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "Register Live Classifier HUD as 8th property on /system/ hub + scaffold follow-on tasks",
      "created_at": "2026-05-06",
      "completed_at": "2026-05-07",
      "path": "/z/public_html/system/index.html",
      "note": "Shipped 2026-05-07. Live Classifier HUD card added to /system/ between Risk Manager HUD and MTF VP Scanner with prod link to /scorecards/james-live.html + watchlist JSON link. Follow-on tasks already filed: roberto-classifier-refresh-cron (P1), roberto-classifier-overrides-persist (P2), live-classifier-engine (completed 5/5).",
      "property": "live_classifier",
      "project": "atlas"
    },
    {
      "id": "catalog-refactor-shared-json",
      "phase": "phase-0",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Catalog refactor: extract 16-strategy catalog to /strategies/catalog.json (one source for 5 surfaces)",
      "created_at": "2026-05-01",
      "note": "Bumped 5+ consecutive days. _scientific_overlay.js, /strategies/, /sizing/, playbooks.json, and now classify/*.json all inline the same catalog. Extract to one /strategies/catalog.json fetched by all surfaces. Stops drift before it compounds.",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "mgp-recap-stability",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Investigate recurring MGP recap-edition silent failures (4/28 + 4/29 + 5/1 missed)",
      "created_at": "2026-05-01",
      "note": "Three of five trading days had no recap edition. Check gen_mgp_intraday.py recap branch + scheduled-task firing time + last successful recap log. 5/5 ran clean; confirm regression genuinely closed before deprioritizing.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "register-daily-stop-system-hub",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Register /daily-stop/ tool on /system/ hub (currently missing from properties list)",
      "created_at": "2026-05-06",
      "completed_at": "2026-05-07",
      "path": "/z/public_html/system/index.html",
      "note": "Shipped 2026-05-07. Daily Stop tool card added to /system/ hub. Cockpit role tag (Sizer), prod link to /daily-stop/ + Sizing Calc link. Notes the user_settings.php storage backend, sibling Sizing Calc, and the outstanding mgp-daily-stop-pipe wiring task.",
      "property": "daily_stop",
      "project": "atlas"
    },
    {
      "id": "locate-bitcoin-todos",
      "phase": "phase-hold",
      "status": "pending",
      "owner": "user",
      "priority": "P2",
      "title": "Locate Bitcoin/IBIT TODO doc on Z drive (location unknown)",
      "created_at": "2026-05-06",
      "note": "James recalls writing a Bitcoin-related TODO list somewhere on Z drive. 2026-05-06 EXHAUSTIVE SEARCH (filename + content) of /trading/a1p_journal (incl. cursor/, src/), /gt_dev, /setup_daily_reporting, /public_html, scorecard_pipeline. All hits are legitimate non-TODO files: KITE backtest summary HTMLs (BTCUSD as test symbol), Vital Dawn email captures, IBIT win-data assets, monthly report HTMLs (incl. unexplained 'james-domino' subdir), Symfony vendor noise. Conclusion: no standalone TODO/planning doc exists at any reachable path. Need user hint: filename pattern, rough date written, directory guess, OR confirmation that it lives off-Z (laptop, Notion, email draft).",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "cron-update-freshness",
      "phase": "phase-0",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Cron /system/update_freshness.py to keep freshness strip auto-refreshing",
      "path": "/z/public_html/system/update_freshness.py",
      "created_at": "2026-05-06",
      "note": "Script ships ready. Suggested cron: */30 9-17 * * 1-5 python /z/public_html/system/update_freshness.py --root /z/public_html (every 30 min during trading hours). Closes the silent-regression loop permanently. ~2 min of work.",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "phase-4-trader-of-the-week",
      "phase": "phase-4",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Trader-of-the-Week scorecard review session (rotating: Nolan / Preston / Brian / Thomas)",
      "created_at": "2026-05-06",
      "note": "Weekly cadence. Pick a trade day, review it with one of the 4 internal traders (rotate through: Nolan -> Preston -> Brian -> Thomas). Output: annotated /playbooks/ entry showing the trade through their lens. Two outputs per session: (1) teaching content for /playbooks/, (2) trader feedback on what's confusing / missing on the scorecard. Doesn't require multi-trader scorecard infra to start \u00e2\u20ac\u201d works against James's existing scorecards. First session sets the template.",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "phase-4-trader-selector",
      "phase": "phase-4",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Trader selector dropdown on scorecard index (driven by ?user= URL param)",
      "created_at": "2026-05-06",
      "note": "Companion to multi-trader-scorecards. Once gen_scorecard.py is parameterized by --user, surface a trader picker on /scorecards/ index. Reads from main.users (joined with main.journal to skip empty traders). Same dropdown component will be reused on monthly + accountability \u00e2\u20ac\u201d see phase-2_5/shared-user-dropdown.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "phase-4-trader-feedback-loop",
      "phase": "phase-4",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "First-scorecard feedback session with each of the 4 traders",
      "created_at": "2026-05-06",
      "note": "After multi-trader-scorecards ships: schedule 4 short sessions (one per trader). Each trader sees their first scorecard, reads the journal/grading + classifier output, gives feedback on what's confusing or missing. Concrete output: a punch list of UX edits + missing fields. Drives the next iteration. Owner=user because James drives the conversation; claude implements the fixes that come out of it.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "alex-range-break-rules",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "alex",
      "priority": "P1",
      "title": "Codify Range-Break strategy rules from Alex's Notion (entry / stop / target / best-when)",
      "created_at": "2026-05-06",
      "note": "Alex sent the Range Break Notion. First step in onboarding him to Strategy Library. Translate the Notion rules into the same JSON shape used by the 16 catalog entries on /strategies/ (id, family, direction, thesis, entry_sd, stop_sd, target_sd, gates[], best_when, alias). Reference: existing 'Backside Long' canonical entry as the template. Output: PR-ready strategy JSON ready for catalog merge.",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "alex-range-break-backtest",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "alex",
      "priority": "P1",
      "title": "Range-Break: walk-forward backtest published at /backtests/range-break/",
      "created_at": "2026-05-06",
      "note": "Once rules codified (alex-range-break-rules). Same walk-forward framework as fl2-band-trail and vwap-pull-in. Reuses gen_walkforward.py. Publishes: IS/OOS Spearman, deflated Sharpe, per-month equity curve. Output gives Alex's strategy a verdict tag (PROVEN / WORKING / MARGINAL / NEGATIVE) for the registry.",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "alex-range-break-register",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "alex",
      "priority": "P1",
      "title": "Register Range-Break in /strategies/ Strategy Library + canonical catalog",
      "created_at": "2026-05-06",
      "note": "Final step. Add range-break card to /strategies/ index.html + entry to the (about-to-be-extracted) catalog.json (see catalog-refactor-shared-json \u00e2\u20ac\u201d order matters: do that refactor first so Alex only writes JSON, not HTML). Card includes the SVG band diagram + edge stats from the walk-forward.",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "roberto-classifier-refresh-cron",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "roberto",
      "priority": "P1",
      "title": "Live Classifier HUD: auto-refresh classify/<SYMBOL>.json during market hours",
      "created_at": "2026-05-06",
      "note": "James-live HUD shipped 5/5 but the classify/*.json files are static. Wire a refresh job: every 1-min during RTH, walk watchlist, pull live SD position from indicator_alerts, recompute the 6-playbook scoring, write JSON. james-live.html already polls; just need the producer side. Likely: extend gen_scorecard.py classifier or new tiny script /scorecards/classify/refresh.py.",
      "property": "live_classifier",
      "project": "classifier"
    },
    {
      "id": "roberto-classifier-overrides-persist",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "roberto",
      "priority": "P2",
      "title": "Persist Live Classifier manual overrides to DB (currently localStorage only)",
      "created_at": "2026-05-06",
      "note": "strategy_tag_overrides.json is the current state \u00e2\u20ac\u201d flat file, single user. Promote to a small DB table keyed (user_id, symbol, trade_date) so overrides survive across sessions and become per-trader. Mirrors the journal-persistence work for scorecards.",
      "property": "live_classifier",
      "project": "classifier"
    },
    {
      "id": "mtf-vp-tradingview-handler",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "MTF VP Scanner: TradingView ticket-in handler (Path B from project memo)",
      "created_at": "2026-05-06",
      "note": "Per memory project_mtf_scan_engine.md, Path A (on-demand worker queue) shipped 4/28. Path B is the TradingView webhook entry point: a Pine alert fires, hits an endpoint, queues a scan, posts the result. Endpoint already mostly stubbed; needs the handler + auth + idempotency. (Originally mis-assigned to Roberto on 2026-05-06; James clarified Roberto's actual work is at https://claude.ai/public/artifacts/1b9da8d6-6627-4f4b-b325-2c65a33bf4cc \u00e2\u20ac\u201d see roberto-onboarding-from-artifact.)",
      "property": "mtf_vp_scanner",
      "project": "mtfvp"
    },
    {
      "id": "roberto-onboarding-from-artifact",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Review Roberto's Claude artifact + slot his real work into the right property",
      "created_at": "2026-05-06",
      "note": "Roberto sent his Claude work at https://claude.ai/public/artifacts/1b9da8d6-6627-4f4b-b325-2c65a33bf4cc (claude.ai artifacts are 403 from WebFetch \u00e2\u20ac\u201d needs a James-side pass). NOT MTF VP Scanner per James's correction. Once reviewed: (1) identify which property the artifact's theme belongs to, (2) replace this task with concrete onboarding tickets in that property, (3) confirm Roberto wants to keep collaborating and pick first ticket. Roberto is ongoing \u00e2\u20ac\u201d treat as recurring collaborator, not one-off.",
      "property": "pm_system",
      "project": "classifier"
    },
    {
      "id": "system-hub-13-properties",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Update /system/ hub diagram to show all 13 properties (currently shows 7-property pipeline only)",
      "created_at": "2026-05-06",
      "completed_at": "2026-05-07",
      "path": "/z/public_html/system/index.html",
      "note": "Shipped 2026-05-07. Header bumped from 'The 8 properties' to 'The 13 properties'. Cards added: Live Classifier HUD, Daily Stop tool. Existing cards retained: Accountability HUD v2, Scorecards, Monthly Reports, Param Surface, Playbooks, Backtests + Walk-Forward, MGP, Risk Manager HUD, MTF VP Scanner, Strategy Library, User Strategy Upload (proposed). Foundation and PM System remain cross-cutting (no card; covered by /todos/ cockpit + freshness panel).",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "freshness-db-source-side",
      "phase": "phase-0",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Freshness dashboard DB-source half: populate sources.* in /system/freshness.json from main.journal + indicator_alerts last-row timestamps",
      "created_at": "2026-05-06",
      "note": "/system/update_freshness.py currently fills only the property side (file mtimes). The sources block stays as 'needs_db_tunnel'. Add a sibling Python pass that opens the SSH tunnel (or runs on a2) and queries: SELECT MAX(trade_date) + COUNT(*) WHERE trade_date >= today-7 from main.journal + alerts_tradingview.indicator_alerts + main.daily_report_cards. Writes the full sources block. Together with property side, this catches BOTH silent-regression patterns (file not regenerated AND ingestion stopped).",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "freshness-smb-cross-host",
      "phase": "phase-0",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Freshness dashboard cross-host smb mtimes (accountability + monthly + param_surface)",
      "created_at": "2026-05-06",
      "note": "Three properties live on smb.a1prospect.com so update_freshness.py running on trsportsacademy can't see their mtimes. Options: (a) sibling update_freshness.py on the smb host writes to a shared JSON merged at fetch time, (b) HTTP HEAD probes against each property URL using Last-Modified header, (c) endpoint on smb that returns the mtime block. Option (b) is cheapest \u00e2\u20ac\u201d just add a function that does HEAD + parses Last-Modified.",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "cockpit-multi-collaborator-owners",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Cockpit: extend owner-split bar to handle 7 owners (jimbo / alex / roberto / 4 internal traders) instead of just user/claude",
      "created_at": "2026-05-06",
      "completed_at": "2026-05-07",
      "path": "/z/public_html/todos/index.html",
      "note": "Shipped 2026-05-07. Velocity owner-split bar now renders one segment per owner present in the data, with 9 distinct colors (claude blue, James purple, Jim gold, Alex green, Roberto orange, Nolan teal, Preston pink, Brian deep-purple, Thomas amber). Filter chip strip auto-populates per-owner chips via renderDynamicOwnerChips() \u00e2\u20ac\u201d any owner that has open tasks gets a clickable button with count. data-filter='owner:<name>' pattern handled in filterPredicate. Ordered fixed (jimbo, alex, roberto, traders) so layout is stable.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "phase-4-per-trader-playbook-focus",
      "phase": "phase-4",
      "status": "pending",
      "owner": "user",
      "priority": "P2",
      "title": "Per-trader playbook specialization (each of Nolan/Preston/Brian/Thomas picks a primary setup)",
      "created_at": "2026-05-06",
      "note": "Each of the 4 internal traders picks one canonical setup to specialize in (e.g. Nolan = Backside Long, Preston = VWAP Pull-In, etc.). Becomes the basis for: (1) personalized scorecard grading weights, (2) per-trader playbook content focus, (3) trader-of-the-week rotation theme. Drives why their multi-trader scorecard differs from yours. Output: a small registry table (trader -> primary setup) referenced by gen_scorecard.py.",
      "property": "playbooks",
      "project": "multi_trader"
    },
    {
      "id": "monthly-per-trader-view",
      "phase": "phase-4",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Per-trader monthly report (extend /james-monthly-reports/ to support ?user= param)",
      "created_at": "2026-05-06",
      "note": "Companion to phase-4-trader-selector. Monthly Dashboard v2 (gen_monthly_dashboard.py) currently hardcoded to James. Parameterize by --user, namespace output (e.g. /james-monthly-reports/ -> /monthly/<trader>/). Each trader gets their own equity curve / by-symbol breakdown / Sortino + Calmar ratios. Depends on multi-trader-scorecards being parameterized first.",
      "property": "monthly_reports",
      "project": "multi_trader"
    },
    {
      "id": "acct-v2-survey-baseline",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "claude",
      "title": "Survey accountability-v2 dev environment + add to system tracking",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/",
      "completed_at": "2026-05-07",
      "created_at": "2026-05-07",
      "note": "Audited dev tree on 2026-05-07. Found: README.md + BUILT.md (2026-04-24 baseline) + docs/ (ARCHITECTURE, HANDOFF-NOW, PROMOTE-DEV-TO-PROD, etc), 21 SQL migrations (010-021 newer than BUILT.md), 4 Node services (position-projector, pnl-engine, fills-poller, marks-fallback), 17 PHP endpoints, HUD + hud-orderlab sub-system, kore-strategies/AccountabilityTickBridge.py, scripts/ for backfill+reconcile+deploy. Per BUILT.md as of 4/24: schema + projection + 2 services + 6 PHP endpoints + HUD all live. Live URLs: v2 HUD https://smb.a1prospect.com/live-accountability-v2/hud/ , v2 health https://smb.a1prospect.com/live-accountability-v2/api/health.php , v1 HUD https://smb.a1prospect.com/live-accountability/ (preserved for rollback).",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-services-task-scheduler",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P1",
      "title": "Move accountability-v2 Node services from nohup \u00e2\u2020\u2019 Windows Task Scheduler (production-grade)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/services/",
      "created_at": "2026-05-07",
      "note": "Per BUILT.md 'Where to go next' #1: services (position-projector, pnl-engine, fills-poller, marks-fallback) are running via nohup/PowerShell background. Add Task Scheduler entries beside the existing v1 daemons (FL2_EOD_Reports template). Survives reboots. Jimbo's lane \u00e2\u20ac\u201d he runs the live infra.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-fills-poller-bridge",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P1",
      "title": "Wire v1 fills-poller to also POST into v2 /api/ingest.php (dual-write transition)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/ingest.php",
      "created_at": "2026-05-07",
      "note": "Per BUILT.md 'Where to go next' #2: today's fills already arrive into alerts_tradingview.fills via v1's poller; v2's position-projector picks them up via LISTEN/NOTIFY. But for double-safety during transition, have v1 also call ingest.php directly. No new code on v1, just a curl/HTTP shim. Confirms parity before deprecating v1.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-kore-tick-bridge-forward-test",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Upload AccountabilityTickBridge.py to KITE forward-test + verify ticks flow into marks_live",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/kore-strategies/AccountabilityTickBridge.py",
      "created_at": "2026-05-07",
      "note": "Per BUILT.md 'Where to go next' #3: ktg.Kite().strategy_upload() the bridge. Then watch marks_live table fill in for the dynamic symbol set returned by /api/bridge-symbols.php. James-side because it requires KITE upload credentials.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-positions-history-seed",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P2",
      "title": "Seed positions_history from v1 EOD state so swing/carry attribution is right",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/sql/",
      "created_at": "2026-05-07",
      "note": "Per BUILT.md 'Where to go next' #4: without yesterday's positions_history seeded, qty_carried + carry_basis show as zero on day-1 carry trades. Backfill script: read v1 data.json's last EOD snapshot, INSERT INTO positions_history. One-shot.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-retire-v1-marks-fast-daemon",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P2",
      "title": "Retire v1 marks-fast-daemon once v2 Kore tick bridge is verified live",
      "created_at": "2026-05-07",
      "note": "Per BUILT.md 'Where to go next' #5. Sequenced AFTER acct-v2-kore-tick-bridge-forward-test confirms marks_live is populating from Kore. Then stop v1's marks daemon \u00e2\u2020\u2019 less infra to babysit, fewer race surfaces.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-recent-migrations-doc-delta",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P2",
      "title": "Update BUILT.md or write delta doc \u00e2\u20ac\u201d migrations 010-021 added since 4/24 baseline",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/BUILT.md",
      "created_at": "2026-05-07",
      "note": "BUILT.md is dated 2026-04-24. 11 new migrations have landed since (daily_settings, positions_view_use_kti_netpnl, position_snapshots, rolling_commission_estimate, historical_breach_from_effective_settings, cto_and_running_pnl, pin_backfilled_cto_to_trigger_trade, carry_forward_effective_settings, today_views_use_effective_settings, breach_uses_gross_pnl, trade_action_requests + undo). Either bump BUILT.md or add CHANGELOG.md so future Claude/teammates can catch up without reading every migration.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-trade-action-requests-feature",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Trade Action Requests feature \u00e2\u20ac\u201d finalize (close/stop intent flow: hud-orderlab + migration 021 + trade-action.php)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/sql/021_trade_action_requests.sql",
      "created_at": "2026-05-07",
      "note": "Per hud-orderlab survey 2026-05-07: this feature is a 3-piece set \u00e2\u20ac\u201d (1) SQL migration 021_trade_action_requests.sql defines the table, (2) api/trade-action.php endpoint inserts into it (returns 503 if migration not applied), (3) hud-orderlab/ is a sandbox copy of hud/ where the close/stop intent UI is being prototyped via trade-action-modal.js. Optional webhook config keys: trade_action_webhook_url, trade_action_webhook_enabled. To finalize: confirm migration 021 is applied to dev DB, link hud-orderlab from main nav (or fold modal into main hud/), test end-to-end. To abandon: run 021_trade_action_requests_undo.sql + delete hud-orderlab/ + delete api/trade-action.php (cleanup checklist in hud-orderlab/README.md).",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-hud-orderlab-survey",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "claude",
      "title": "Survey hud-orderlab/ sub-system \u00e2\u20ac\u201d document what it is, decide if it needs its own property",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/hud-orderlab/",
      "created_at": "2026-05-07",
      "completed_at": "2026-05-07",
      "note": "Surveyed 2026-05-07. RESULT: hud-orderlab is NOT a separate sub-system \u00e2\u20ac\u201d its README is explicit: 'a copy of ../hud/ used to try close/stop intent flows without changing the main Accountability HUD.' Sandbox UI for the trade-action-requests feature. Files: index.html + js/ (10 files including trade-action-modal.js) + css/ + config/ (env.dev.json, env.json, env.prod.json). Backed by migration 021_trade_action_requests.sql + api/trade-action.php endpoint. Returns 503 migration_required if table missing. README has a cleanup checklist for deletion if the experiment is abandoned. CONCLUSION: stays under property=accountability_hud. No new property needed. The acct-v2-trade-action-requests-feature task captures the related work.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-promote-dev-to-prod",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Execute dev\u00e2\u2020\u2019prod promotion sequence per docs/PROMOTE-DEV-TO-PROD.md",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/docs/PROMOTE-DEV-TO-PROD.md",
      "created_at": "2026-05-07",
      "note": "Doc PROMOTE-DEV-TO-PROD.md exists in dev/docs \u00e2\u20ac\u201d implies promotion isn't a one-step copy but a sequenced procedure. Read the doc, schedule the promotion window, follow steps. Prerequisite: services running on Task Scheduler (acct-v2-services-task-scheduler) + Kore bridge verified (acct-v2-kore-tick-bridge-forward-test) + migrations doc current (acct-v2-recent-migrations-doc-delta).",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-bug-risk-bump-ignores-date-range",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P0",
      "title": "BUG: Risk Bump Progress bar does not respect the selected date range (always shows 5-day rolling, even when range = full month)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/trader-range.php",
      "created_at": "2026-05-07",
      "note": "Reported by James 2026-05-07 with two screenshots from the live DEV HUD (https://smb.a1prospect.com/live-accountability-v2-dev/hud/?t=james&range=c...). REPRO: select James + date range 2026-04-01 \u00e2\u2020\u2019 2026-05-08. OBSERVED: Today's P&L tile correctly reflects the range (+$467.66 across 3328 round trips, 45 swing exits, best day +$177.78, worst MDD -$227.96). But the 'RISK BUMP PROGRESS (5-DAY ROLLING)' bar shows -$27.74 / +$500.00 with '4 ACTIVE DAYS' and '-5.5% toward bump target' \u00e2\u20ac\u201d the 4-day count gives away that the bar is stuck on a fixed last-N-days window, NOT the 38-day selection. EXPECTED: when a multi-week range is selected, the bump-progress bar should either (a) integrate bump deltas across the full selected range, or (b) be hidden / clearly labeled 'always rolling 5d from today, ignores filter'. SUSPECTED CODE: api/trader-range.php (date params not flowing through to bump calc) or hud/js/render.js (rendering a always-current-window value inside a section that visually belongs to the range filter). DML/CVaR derived stop = $130 per day. UI ALSO shows Range Progress Toward 2:1 Target correctly green (gray over the green = past 1:1 stop midpoint) so that bar IS range-aware \u00e2\u20ac\u201d confirming this is a per-component bug, not a global filter wiring problem. SCREENSHOTS: James to drop into /todos/issues/ (dir created 2026-05-07) when convenient; describing in this note for now. ACCOUNTABILITY CODE OFF-LIMITS to Claude per feedback_hud_off_limits.md \u00e2\u20ac\u201d Jim must approve and execute the fix.",
      "property": "accountability_hud",
      "project": "acct_v2",
      "sprint": 1
    },
    {
      "id": "pm-nightly-silent-failure-detection",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Nightly PM SKILL: detect + alert on silent failures (5/7 23:04 ET fire produced no output)",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\nightly-project-manager\\SKILL.md",
      "created_at": "2026-05-08",
      "note": "DIAGNOSIS 2026-05-08: nightly-project-manager scheduled task DID fire 2026-05-07 23:04 ET (lastRunAt=2026-05-08T03:04:56.959Z UTC per scheduled-tasks list_scheduled_tasks). But it produced no /todos/daily/2026-05-07.md and didn't bump meta.last_nightly_run. Pure silent failure. Most likely: WebDAV scan timeout during Step 1 (the SKILL itself flags 'Whole-tree recursive find from /z/public_html can stall on WebDAV \u00e2\u20ac\u201d if the public_html scan exceeds 60s, fall back to a per-subdir loop'). Either the fallback didn't trigger, or a later step (Step 2/3 write) failed without surfacing. FIX: (a) every step writes a heartbeat to /z/logs/nightly-<date>.log with start/end timestamps so we can see WHERE it bailed, (b) at Step 6, if no .md was produced, post a stale-flag to /system/freshness.json + send a notify.log alert, (c) wrap each Find/find call in a 30s timeout with retry. Backfilled 2026-05-07.md manually 2026-05-08 morning to close the immediate gap. SECOND PROBLEM (added 2026-05-08): James reports he's been manually accepting permission prompts on the routine's writes \u00e2\u20ac\u201d defeats the autonomy of a scheduled job. That's a separate path-pattern issue tracked in claude-code-webdav-path-workaround; the symlink fix there should let this routine run unattended once it's the project's working directory. Both issues need to land before this routine is reliable. UPDATE 2026-05-11 (morning catch-up fire): Part (a) heartbeat is now PROVEN to work \u00e2\u20ac\u201d 7:43 ET fire wrote LAUNCHED but never DONE, giving us first concrete evidence of the agent-crash-mid-run failure mode (distinct from agent-never-launched). Sequence in /z/logs/nightly-startup.log: 5/10 23:10 ET manual-proxy LAUNCHED+DONE \u00e2\u2020\u2019 5/11 07:43 ET daemon LAUNCHED (no DONE = crashed mid-run) \u00e2\u2020\u2019 5/11 08:05 ET daemon LAUNCHED again (this run). Three open follow-ups: (1) parts (b) and (c) of the fix \u00e2\u20ac\u201d auto stale-flag + per-step timeouts \u00e2\u20ac\u201d are still pending; heartbeat alone tells us THAT it failed, not WHY. (2) Root-cause the 07:43 ET crash: pull its transcript JSONL from ~/.claude/projects/<sanitized-cwd>/ to identify the last successful step (could be path-pattern permission prompt timeout, WebDAV stall, SKILL parse error). (3) Host-sleep mitigation: cron fires at 23:00 ET but machine often asleep at that hour; either wake-on-schedule, move cron earlier, or accept catch-up-on-wake as normal.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "pm-nightly-refactor-for-loop-scan",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Refactor nightly PM SKILL: replace 'for d in ...; do ls ...; done' fallback with Python helper so it stops triggering Bash permission prompts",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\nightly-project-manager\\SKILL.md",
      "created_at": "2026-05-10",
      "completed_at": "2026-05-10",
      "note": "SHIPPED 2026-05-10 (three passes): (PASS 1) wrote helper at Z:\\james\\.claude\\scripts\\scan_recent_activity.py \u00e2\u20ac\u201d pure-Python implementation of the original ls+awk logic, single 'py ...' invocation matches Bash(py *). Fixed dir list bug: SKILL had 'db_reports' which doesn't exist under public_html, removed it and added 'daily-stop' + 'backtests'. Updated SKILL.md Step 1 fallback section to call the helper. (PASS 2) Discovered .claude/ paths trigger Claude Code's 'sensitive file' warning on edits \u00e2\u20ac\u201d separate prompt bucket from the bash-allowlist and path-pattern issues. RELOCATED helper to Z:\\setup_daily_reporting\\reports\\scan_recent_activity.py (alongside other nightly helpers like update_todos_for_nightly.py and render_daily.py). Deleted old .claude/scripts/ location. (PASS 3 \u00e2\u20ac\u201d Option B): Per James's request, made the helper Step 1's PRIMARY scan path instead of just the fallback. New design uses TWO calls: Call A scans the 3 small roots recursively (~50s); Call B does top-level subdir mode over public_html (~10s). Total ~60s on WebDAV \u00e2\u20ac\u201d comparable to the old find primary but without the stall risk. Full recursive over public_html was tested and stalls on WebDAV after 5+ min (too many files), so subdir mode is the right shape for that root specifically. Also added bash-path normalization to the helper (/z/foo \u00e2\u2020\u2019 Z:/foo) so paths from Git Bash work in Python's os.path.isdir. CONTEXT: James caught the autonomous nightly pausing on a permission prompt 2026-05-10 \u00e2\u20ac\u201d leading-token 'for' + simple_expansion couldn't be allowlisted. RELATED still pending: claude-code-webdav-path-workaround (path-pattern blocker on file writes, separate bucket); pm-nightly-silent-failure-detection (silent failure heartbeats). All three needed for fully reliable unattended fires.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "claude-code-webdav-path-workaround",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Test mklink workaround for Claude Code 'suspicious path pattern' prompt on WebDAV-backed projects (unblocks nightly PM routine + agentic sessions)",
      "created_at": "2026-05-08",
      "note": "Claude Code's safety layer flags suspicious Windows path patterns and forces manual approval on EVERY file write \u00e2\u20ac\u201d even with 'Allow bypass permissions mode' enabled \u00e2\u20ac\u201d because the WebDAV mount point name leaks into the canonical project path: C:\\Users\\jtdom\\.claude\\projects\\--a1prospect-com-SSL-2078-DavWWWRoot-james--claude\\. The '--' prefix (looks like a CLI flag), the 'SSL-2078-DavWWWRoot' segment (looks like a network/server path), and overall character density match the path matcher's heuristics. Result: 20+ file edits per session = 20+ approval clicks, which kills agentic workflow on strategy registry, Pine Script iteration, and DRC automation. NIGHTLY PM ROUTINE IMPACT (added 2026-05-08 by James): the scheduled nightly-project-manager fires writes against todos.json + /todos/daily/<date>.md and James has been manually accepting permission prompts on those writes. That defeats the whole point of an autonomous nightly run \u00e2\u20ac\u201d a scheduled job by definition isn't supervised. This workaround, if it works, unblocks: (a) any agentic session in the project; (b) the nightly PM routine running unattended; (c) any future Claude scheduled-task that touches files under /z/. PROPOSED FIX: create a Windows directory symlink from a clean path (e.g. C:\\dev\\a1prospect) to the real WebDAV-backed project folder. If Claude Code reads the clean path without resolving back to the canonical UNC, the matcher won't trigger. CAVEAT: not yet verified \u00e2\u20ac\u201d Claude Code may resolve symlinks to canonical paths, in which case this approach won't help. TEST-FIRST PROTOCOL: (1) verify symlinks work via throwaway test (mkdir C:\\test_target; mklink /D C:\\test_link C:\\test_target; dir C:\\test_link; rmdir C:\\test_link [NO /s flag] then rmdir /s C:\\test_target). (2) xcopy backup the real project to C:\\backup\\a1prospect_backup. (3) Admin cmd: mkdir C:\\dev; mklink /D \"C:\\dev\\a1prospect\" \"C:\\Users\\jtdom\\.claude\\projects\\--a1prospect-com-SSL-2078-DavWWWRoot-james--claude\". (4) dir C:\\dev\\a1prospect to verify. (5a) Open project in Claude Code via the new clean path, try a file edit, watch for the safety prompt. (5b) Repoint the nightly-project-manager scheduled task at the clean path and let it fire once unattended \u00e2\u20ac\u201d verify no prompts queued + writes complete. If prompt stops \u00e2\u2020\u2019 roll out per project on BOTH machines (symlinks are local, not synced) and update the scheduled task working directory on each. If prompt still fires \u00e2\u2020\u2019 rmdir C:\\dev\\a1prospect (no /s) and fall back to Plan B: maintain working copy on C:\\dev with a sync-to-Z scheduled task so Claude Code never sees the WebDAV path. SAFETY: mklink only creates a pointer, doesn't move/modify files. Cross-device sync via Z: WebDAV stays intact. NEVER use 'rmdir /s' on a symlink \u00e2\u20ac\u201d it can recursively delete the target. 'Allow bypass permissions mode' toggle stays ON (this fix is for the secondary path-pattern check). RELATED: pm-nightly-silent-failure-detection (different bug \u00e2\u20ac\u201d silent failure heartbeats \u00e2\u20ac\u201d but same routine). Source: writeup pasted into chat by James 2026-05-08; ChatGPT-drafted but James-reviewed.",
      "property": "pm_system",
      "project": "atlas"
    },
    {
      "id": "acct-v2-hud-trader-state-card",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P2",
      "title": "Add Trader State card (mood meter / OVERREACTION ZONE) to Accountability HUD as bigger-picture summary; drill-through to Execution Scorecard",
      "path": "/z/public_html/scorecards/trade_scorecard_2026-04-10.html",
      "created_at": "2026-05-08",
      "note": "James asked by name 2026-05-08 to surface Trader State on the Accountability HUD as bigger-picture, with full detail remaining on the Execution Scorecard. SOURCE MODULE: section 'TRADER STATE / MOOD METER' in trade_scorecard_*.html (around line 546). Renders four bands (Opportunity / Hesitation / Overconfidence / Overreaction) as horizontal bars with %, plus 'BEHAVIORAL SIGNALS DRIVING THIS READ' table: ticket count vs baseline, shares traded, recent streak, mixed-bias entries, revenge re-entries, consecutive losers, tight loss clusters, risk acceptance, daily stop usage. Top pill shows current zone (e.g. OVERREACTION ZONE with 66% bar). HUD CARD SPEC: per-trader summary tile showing current zone pill + 1-2 strongest behavioral triggers; click drills to that trader's daily Execution Scorecard. Don't overwhelm the HUD \u00e2\u20ac\u201d show the read, not the full signal table. CONSTRAINT: feedback_hud_off_limits.md \u00e2\u20ac\u201d accountability code (live-accountability-v2) is OFF-LIMITS to Claude; James/Jim implements. Claude can spec, design the card layout, and provide reference screenshots from the existing scorecard module. RELATED: acct-v2-hud-composure-trend-card, acct-v2-hud-strategies-mood-card.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-hud-composure-trend-card",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P2",
      "title": "Add Composure Trend card (trailing 30d, 20 sessions) to Accountability HUD; drill-through to Execution Scorecard",
      "path": "/z/public_html/scorecards/trade_scorecard_2026-04-14.html",
      "created_at": "2026-05-08",
      "note": "James asked by name 2026-05-08 to surface Composure Trend on the Accountability HUD. SOURCE MODULE: section 'COMPOSURE TREND' in trade_scorecard_*.html (around line 785). Renders a Daily Composure 0-100 line + Daily Stop Usage % overlay across last ~20 sessions. Header shows TODAY value + 30d AVG + TREND label (Trending up / down). Trend bands: \u00e2\u2030\u00a575 FOCUSED, 55-74 STEADY, 40-54 WATCH IT, 25-39 DRIFTING, <25 OFF PROCESS. HUD CARD SPEC: per-trader compact tile with today's composure value + 30d avg + trend arrow + tiny sparkline; click drills to that trader's daily Execution Scorecard for the full chart and band overlays. Keep the card narrow \u00e2\u20ac\u201d meant to be skimmed alongside the existing P&L / stop% tiles. CONSTRAINT: feedback_hud_off_limits.md \u00e2\u20ac\u201d Claude can spec, Jim implements. RELATED: acct-v2-hud-trader-state-card, acct-v2-hud-strategies-mood-card.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "acct-v2-hud-strategies-mood-card",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P2",
      "title": "Update strategies/per-symbol mood sentiment and surface a condensed view on Accountability HUD",
      "path": "/z/public_html/scorecards/mood-meter-preview.html",
      "created_at": "2026-05-08",
      "note": "James asked by name 2026-05-08 to update the strategies mood sentiment AND get a condensed view onto the Accountability HUD alongside Trader State + Composure Trend. SOURCE MODULE: 'Per-Symbol Mood \u00c3\u2014 Tickets' on Execution Scorecard \u00e2\u20ac\u201d two gauges per symbol (Tickets churn 0/100 vs per-symbol median, Mood 0/100 avg composure across that symbol's entries), sorted worst-mood first with badges OFF PROCESS / DRIFTING / WATCH IT / STEADY / FOCUSED. Existing files: /z/public_html/scorecards/mood-meter-preview.html, mood-ticket-widget.html. HUD CARD SPEC: per-trader, show top 3 worst-mood symbols today as small chips (symbol + mood badge + ticket count). Click drills to that trader's Execution Scorecard for the full per-symbol grid. OPEN QUESTION: did James intend ALSO to update the Execution Scorecard widget itself (recompute ranges, refresh thresholds, restyle) \u00e2\u20ac\u201d or only to bring it onto the HUD? Verify scope with Jim before Jim builds. CONSTRAINT: feedback_hud_off_limits.md \u00e2\u20ac\u201d Claude can spec the HUD card; Jim implements both the Execution Scorecard update and the HUD card. RELATED: acct-v2-hud-trader-state-card, acct-v2-hud-composure-trend-card.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "risk-hud-accountability-hud-dedup-review",
      "phase": "phase-0",
      "status": "pending",
      "owner": "user",
      "priority": "P1",
      "title": "Decide whether Risk Manager HUD (/accountability/risk.html, mock) is redundant with Accountability HUD (/live-accountability-v2/hud/) \u00e2\u20ac\u201d merge or retire",
      "path": "/z/public_html/accountability/risk.html",
      "created_at": "2026-05-08",
      "note": "James flagged 2026-05-08 to constantly scan for duplication. The Risk Manager HUD card on /system/ (built 2026-04-21, mock data, 'awaiting live API endpoint from smb subdomain' per risk-hud-api-wire) is described as: stop-utilization gauge per trader, summary row (clear/warning/danger/breach), Risk View sort worst-first, 60s auto-refresh, 4-tier color (green <50%, yellow 50-80%, orange 80-99%, red \u00e2\u2030\u00a5100%). The live Accountability HUD (/live-accountability-v2/hud/) ALREADY shows per-trader stop multiple (Brian -0.98x, Enrique 0.58x, James -0.65x, Nolan 0.95x, Preston 0.00x, Thomas 0.41x as of 2026-05-08 12:17 ET screenshot) and per-trader cards in production. OVERLAP: both compute and render per-trader stop utilization; risk_hud adds Risk-Manager-specific affordances (sort worst-first, summary tiles, auto-refresh) that don't yet exist on accountability_hud. DECISION NEEDED before any of risk-hud-stop-gauge / risk-hud-summary-row / risk-hud-sort-risk / risk-hud-auto-refresh / risk-hud-api-wire move forward: (a) RETIRE risk_hud and merge its Risk-Manager-view affordances into accountability_hud as a 'Risk Manager' filter/sort mode; (b) KEEP risk_hud as a separate Risk-Manager-only view layered on top of the accountability_hud data API; (c) keep both with clear distinct purpose (e.g. risk_hud = read-only operations dashboard, accountability_hud = trader self-service). Hub-first principle: pick one source of truth before building the duplicate. Currently 5 risk-hud-* tasks are blocked behind this decision; all of them describe behavior the Accountability HUD could absorb. Recommend (a) so we don't fork the data layer.",
      "property": "risk_hud",
      "project": "acct_v2"
    },
    {
      "id": "tradingview-chart-viewer-swap",
      "phase": "phase-shared-infra",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Shared chart viewer: drop-in replacement with TradingView Advanced Chart widget",
      "path": "/z/public_html/_shared/chart_viewer.js",
      "completed_at": "2026-05-08",
      "created_at": "2026-05-08",
      "note": "Replaces lightweight-charts engine with TradingView Advanced Chart (https://s3.tradingview.com/tv.js). Public contract preserved: any element with data-ticker=SYM is still clickable; auto-injection via /_shared/nav.js still applies, so no consumer page needs editing. New: extended hours session enabled by default, ESC + backdrop-click close, dynamic re-binding for JS-rendered tickers. Timeframes: 2m/5m/15m/30m/65m(=TV 60m)/1D. /scorecards/api/ticker_chart.php endpoint orphaned (kept on disk). Roll-back: mv chart_viewer.js.pre-tradingview.bak chart_viewer.js.",
      "project": "multi_trader"
    },
    {
      "id": "vwap-param-surface-image-refresh-2026-05-08",
      "phase": "phase-param-surface",
      "status": "completed",
      "owner": "claude",
      "title": "VWAP Param Surface: full TOD + per-symbol image refresh (12 symbols, 7 TOD windows)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "completed_at": "2026-05-08",
      "created_at": "2026-05-08",
      "note": "Full regen of param_surface, param_heatmap, tod_surface_*, tod_heatmap_* PNGs across 12 tickers (BE, CAR, ETHU, IBIT, MSFT, NVDA, ORCL, PLTR, SNDK, USO, WULF + all_symbols_combined) and 7 TOD windows (open 0930-1000, early 1000-1030, mid_am 1030-1130, midday 1130-1300, mid_pm 1300-1430, late 1430-1555, full). tod_sd_expansion + tod_comparison summary plots also rebuilt. Static + DB index.html variants both refreshed. Does NOT close p0-param-surface-registry (still pending \u00e2\u20ac\u201d this was a data refresh, not the registry build).",
      "project": "atlas"
    },
    {
      "id": "mgp-stocks-in-play-tuning-2026-05-08",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "MGP pipeline tuning: stocks_in_play.py + gen_mgp_intraday.py pre-market pass",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_intraday.py",
      "completed_at": "2026-05-08",
      "created_at": "2026-05-08",
      "note": "Coordinated pre-market edit (07:48-07:50 ET) to the picks-selection (stocks_in_play.py) and mid-day MGP generator (gen_mgp_intraday.py). Mid-Day MGP for 2026-05-08 went out at 11:38 ET successfully (notify.log confirmed). Recap MGP at 16:38 also clean. 5/7 mid-day + recap re-emissions were also sent at 07:52 (yesterday backfill).",
      "project": "atlas"
    },
    {
      "id": "mgp-premarket-data",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "MGP morning: pull pre-market OHLCV via KTI and surface gap / range / volume per pick (Phase 2 from 2026-05-08 click-to-chart session)",
      "path": "/z/setup_daily_reporting/reports/stocks_in_play.py",
      "created_at": "2026-05-11",
      "completed_at": "2026-05-11",
      "note": "SHIPPED 2026-05-11 evening. fetch_premarket_bars() batched KTI call (kti price query minute, 04:00-09:30 ET window), compute_premarket_stats() returns pm_open/high/low/last/range/volume/gap_pct/range_atr_pct. Wired into build() \u2014 pm fields merged into each pick's stats dict. rank_pick() now factors pm_gap with thresholds |gap|\u22651.5% AND pm_volume\u2265250k \u2192 \u00b11 score depending on direction-alignment. render_html adds a 'Pre-Market' row to each pick card with last/gap/range/vol. Smoke test on today's picks: MFIC pm_gap=+5.32%, INTC pm_vol=18.7M (gap +1.24%) \u2014 all 5 picks returned PM data via batched call. gen_mgp.py passes parsed['date'] as trade_date so backfills hit the correct session. Live on next morning fire (2026-05-12 05:47 ET). Phase 3 (Lightweight-Charts viewer with KTI data) deferred.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-drop-vital-independent-source",
      "phase": "phase-mgp",
      "status": "in_progress",
      "owner": "claude",
      "priority": "P0",
      "title": "Drop Vital Knowledge entirely: build independent ticker selection + intraday refresh so subscription can be cancelled",
      "path": "/z/setup_daily_reporting/reports/morning_tickers.py",
      "created_at": "2026-05-13",
      "note": "Decision 2026-05-13 (James): after the compliance strip, drop Vital entirely and redirect the subscription cost elsewhere. PHASED PLAN: (2a) build morning_tickers.py with pre-market gappers from KTI minute bars + a known universe (SPY+QQQ components ~600 tickers) \u2014 generates our own daily ticker list. (2b) add build_morning_data() that returns parsed-dict shape used by gen_mgp.py + add --source flag (vital|independent). (2c) modify gen_mgp_intraday.py to skip the Vital IMAP fetch \u2014 intraday pages already only show our own analysis (Index Stack, picks classification, etc.). (2d) verify 3+ trading days of clean output from --source=independent, then James cancels Vital subscription + cancels IMAP forwarding rule. PROGRESS 2026-05-13 evening: Phase 2a + 2b + 2c ALL SHIPPED end-to-end. morning_tickers.py scans 200-ticker universe.json via KTI 1-min bars 04:00-09:30 ET, filters |gap|\u22651.5% AND pm_vol\u2265250K, sorts by |gap_pct| then volume. catalysts.py enriches each pick with yfinance earnings flag + most-recent news headline (handles old+new yfinance shapes, 48h staleness filter, 404-safe for ETFs). morning_data.py wraps it into parsed-dict shape that gen_mgp.py expects (date/day_of_week/tickers/earnings_today/econ_events/commodities/futures/overall_bias/bull_tag). gen_mgp.py --source independent verified end-to-end: 200 symbols \u2192 22 gappers \u2192 15 picks rendered, bias=risk-off from weighted SPY/QQQ/IWM/DIA gaps, convergence banner correctly flagged DIVERGENCE vs Bullish Index Stack, catalyst headlines visible on every pick card, ZERO Vital references in 137KB HTML. gen_mgp_intraday.py --source independent verified end-to-end on both midday + recap editions: skips IMAP entirely, uses --date (or today ET), builds minimal parsed dict (date/day_of_week + empty tickers/commentary/etc.), still loads morning picks from DB and classifies via 5-min bars to cutoff (midday picks: 5 classified, recap picks: 4 stopped 1 active), ZERO Vital references on either intraday page (23-24KB each). Bias on intraday left empty for now (renders 'unknown') \u2014 Phase 2d will wire fresh intraday bias from SPY/QQQ/IWM/DIA daily. NOTE: during testing accidentally triggered a real notify.send_mgp_notice for the midday + recap test runs (forgot --dry-run when verifying file output); James may have received two extra emails 2026-05-13 PM with today's date \u2014 these are valid independent-source content but not on the normal cron. NEXT: switch the cron jobs to --source=independent for tomorrow morning's 5:47 ET run + verify 3+ trading days clean before James cancels Vital subscription.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-historical-pages-cleanup",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "james",
      "priority": "P0",
      "title": "Compliance follow-up: delete (or hide) historical MGP pages 4/17 through 5/12 \u2014 they still contain Vital Knowledge text",
      "path": "/z/public_html/mgp/",
      "created_at": "2026-05-13",
      "note": "PENDING JAMES'S APPROVAL. Today's three pages (5/13 morning + midday + recap) are clean. But ~40 historical pages from 2026-04-17 through 2026-05-12 still on the public web at https://trsportsacademy.com/mgp/<date>.html (and -midday/-recap variants for some dates). Each contains: Section 2 'Vital Dawn Snapshot' with their futures bias + commentary cells, Market Commentary blocks with verbatim/paraphrased Vital bullets, catalyst text per pick, in many cases Subscribe CTA. Risk: every day this stays public is a fresh ToS breach record, court can enjoin per ToS clause. THREE OPTIONS (Claude provided recommendations 2026-05-13): (A) DELETE all historical pages + rebuild manifest + rebuild index \u2014 fastest path to zero exposure; SIP picks remain in DB so analysis is recoverable. (B) Regenerate from saved fixtures for dates that have them in /z/setup_daily_reporting/reports/mgp_fixtures/ \u2014 preserves our analysis layer but timestamps drift + not all dates have fixtures. (C) Hide behind HTTP auth \u2014 preserves data but doesn't address Google cache. RECOMMENDED: Option A. EXECUTE STEPS when James approves: (1) ls /z/public_html/mgp/*.html \u2014 list all files (~40-50 expected); (2) confirm with James which to delete; (3) rm the historical files; (4) cd Z:\\setup_daily_reporting\\reports && py -c 'from gen_mgp import write_manifest, update_archive_index, DEFAULT_OUT_DIR; write_manifest(DEFAULT_OUT_DIR); update_archive_index(DEFAULT_OUT_DIR)' \u2014 rebuild manifest + index; (5) verify https://trsportsacademy.com/mgp/ archive only lists today's entries going forward; (6) optionally check Google Search Console for cached pages and request removal.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-vital-knowledge-compliance-strip",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "Compliance: strip Vital Knowledge content from public MGP pages + remove Subscribe CTA",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_intraday.py",
      "created_at": "2026-05-13",
      "completed_at": "2026-05-13",
      "note": "SHIPPED 2026-05-13. Verified zero 'Vital' references on all three pages of today's MGP (/mgp/2026-05-13.html, -midday.html, -recap.html). Changes: (1) gen_mgp_intraday.py \u2014 removed Market Commentary section + New Flags section from both page and email; simplified subtitle to drop 'from Vital {label}'; replaced flagged-count phrasing with neutral text; fixed unsubscribe footer copy. (2) gen_mgp.py \u2014 removed Subscribe CTA block + subscribe-cta CSS from update_archive_index(); kept Domino Brief as our own brand; renamed pre-market source description; renamed [imap] log message. (3) mgp_template.html \u2014 DELETED Section 2 'Vital Dawn Snapshot' entirely (was the biggest breach: displayed Vital's futures bias, futures cells, index bias note, commodities cells); renumbered Sections 3-8 \u2192 2-7; simplified subtitle and footer; removed 'pulled from Vital Dawn's Catalysts block' phrasing from events section; simplified Section 1 description. (4) stocks_in_play.py \u2014 removed CATALYST label + catalyst text from pick cards (kept the significance chip which is our derived classification + the if_then/setup/entry/stop/target which are all our analysis). (5) convergence.py \u2014 renamed banner text from 'Vital Dawn agree/says' to 'futures-implied bias agree/says' (futures bias is computed from factual bp moves). (6) cleared stale __pycache__/convergence.cpython-313.pyc that was caching old text. Independent analysis still intact: Index Stack v2+v3, Long/Short Recommendations, Pre-Market data, ChipHazzard regime, breadth composite, our commodities, AMZN case study + fixture test, calendar nav. Pipeline now operates 'private intake \u2192 publish own analysis only' \u2014 Vital narrative never appears on public pages.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-commodities-repeat-and-prev-button-fix",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Fix repeating text in Commodities section + broken Prev button on calendar nav",
      "path": "/z/setup_daily_reporting/reports/vital_dawn_parser.py",
      "created_at": "2026-05-12",
      "completed_at": "2026-05-12",
      "note": "Two bugs in one. (1) Commodities section was showing identical text under multiple cards (Treasuries and Brent shared one sentence; Gold/silver/Bitcoin shared another). Root cause: parse_commodities() matched whole sentences containing a keyword, but a single Vital Dawn sentence often mentions multiple commodities. FIX: split text on extra boundaries (parens, semicolons, ' while ', ' and '), then for each keyword pick the clause where it appears EARLIEST (most likely to be the subject). Also added bp->% conversion (singletons '240bp'->'2.4%' AND ranges '1-2bp'->'0.01%-0.02%') so the morning section matches the rest of the MGP's unit convention. Verified all 6 commodities now show unique cleaned text. (2) Prev button on calendar nav was disabled when on the latest entry, because /_shared/mgp_cal.js had Prev/Next semantics swapped (Prev->newer, Next->older); chronological convention is Prev=older, Next=newer. FIX: swapped the variable references in the button calls. Both fixes shipped to today's MGP at /mgp/2026-05-12.html. User must hard-refresh browser to pick up new mgp_cal.js (browser cache).",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "case-study-2026-05-12-AMZN",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "First fixture-driven case study: AMZN 2026-05-12 gap-under-5DMA mean-reversion long in Stage 2 markup",
      "path": "/z/setup_daily_reporting/case_studies/2026-05-12_AMZN_gap-under-5dma/",
      "created_at": "2026-05-12",
      "completed_at": "2026-05-12",
      "note": "SHIPPED 2026-05-12. Bottom-up pivot per James's strategic doc: build from REAL TRADE EXAMPLES. Layout: case_studies/2026-05-12_AMZN_gap-under-5dma/{case.md (5.4KB writeup), fixtures/AMZN_daily.json (122 daily bars), expected.json (asserted classifier output)} + tests/test_case_studies.py (auto-discovery harness). Concrete classifier outputs captured for AMZN 5/11 close: state=markup, shannon_stage=stage_2, hh_cleared=true, ma_stack_up=true, tier=long, close $268.99 below 5DMA $272.28 (the gap-under setup), breakout_trigger $251.39 cleared by 2.43 ATR. Test passes 1/1. PATTERN ESTABLISHED for future trades: drop a new case_studies/<date>_<ticker>_<setup>/ dir with case.md + fixtures/ + expected.json \u2014 test harness auto-discovers it, no per-trade code. James's 'OUTCOME' section in case.md left blank for him to fill in post-trade (filled price, MFE/MAE, realized R, lessons). NOTE on paths: initially saved fixture to wrong WebDAV subtree (/dav/.../james/setup_daily_reporting/case_studies/) because Z: drive and \\\\...\\james\\ are separate virtual roots. Moved to correct Z: location. Don't pass /z/ paths via Python pathlib directly \u2014 they don't resolve on Windows; use the explicit \\\\a1prospect.com@SSL@2078\\DavWWWRoot\\ UNC path or Z:\\... drive letter. NEXT: emit behavioral_class field from chiphazzard_v3 (per Q1 of strategic doc), then wire grade+stage into position-sizing module.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "mgp-intraday-technical-read",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Mid-Day + Recap pages: add Technical Read (Index Stack) panel with multi-TF view (Daily v2/v3 side-by-side + 65m + 30m + 15m)",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_intraday.py",
      "created_at": "2026-05-12",
      "completed_at": "2026-05-12",
      "note": "SHIPPED 2026-05-12. Added render_technical_read() helper to gen_mgp_intraday.py \u2014 stacks five panels: Daily v2.2 + Daily v3 side-by-side, then 65m / 30m / 15m using v2.2 (bug fix is daily-only). Each call wrapped in best-effort try/except so a single TF KTI failure doesn't kill the section. Inserted into render_page_html as new H2 'Our Technical Read' right between bias summary card and Market Commentary. Mirrored chiphazzard panel CSS classes (.ch-panel/.ch-head/.ch-grid/.ch-card/.ch-tk/.ch-state/.ch-label/.ch-comp/.ch-note) from mgp_template.html into the intraday page's <style> block. Bonus fix in chiphazzard.render_panel_html: the heading was hard-coded 'Index Stack (1D)' regardless of the panel's timeframe field \u2014 now reads panel.get('timeframe') so the intraday panels label themselves 65m / 30m / 15m correctly. Smoke test on 2026-05-11-midday.html: all 5 panels rendered with correct TF labels and same-page CSS styling matching morning page. Same code path applies to recap edition \u2014 next recap fire will produce identical layout. Page size: 14k \u2192 26k bytes (the 4 extra panels).",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "chiphazzard-v3-regime-fix",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "ChipHazzard daily regime mis-classification: SPY/QQQ/IWM labeled ANTICIPATE UP when they should be MARKUP \u2014 fix Stage 1\u2192Stage 2 promoter (Weinstein base resistance)",
      "path": "/z/setup_daily_reporting/reports/chiphazzard_v3.py",
      "created_at": "2026-05-12",
      "completed_at": "2026-05-12",
      "note": "SHIPPED 2026-05-12. Built chiphazzard_v3.py with classify_state_v3() implementing Weinstein base-resistance Stage 1\u21922 promoter: breakout_level = max(high, 60-bar window, 20-bar offset); breakout_trigger = breakout_level + 0.5\u00d7ATR14; hh_cleared = close > breakout_trigger. Mirror logic for bearish breakdowns. Output dict shape is a superset of v2.2 \u2014 adds shannon_stage (stage_1 / stage_1_emerging / stage_1_late / stage_2 / stage_3 / stage_4), failing_gate (hh_cleared / ll_cleared / ma_stack_up / etc.), breakout_level, breakout_trigger, dist_to_breakout_atrs, hh_cleared, ll_cleared, tier. Wired as sidecar via _safe_chip_panel() in gen_mgp.py \u2192 calls ch3.render_sidebyside(v2_panel, v3_panel) which produces a 2-col grid 'v2.2 (current production)' | 'v3 (proposed)'. Smoke test on /mgp/2026-05-11.html: v2.2 still shows ANTICIPATE UP for SPY/QQQ/IWM (the documented bug); v3 shows MARKUP for all three with breakout-level chips \u2713 cleared (SPY base $697.84 trigger $701.21 close $739.30 \u2014 cleared by ~$38). Intraday risk lanes still use v2.2 since the bug is daily-only. Validation plan: 3 sessions of side-by-side review, then drop v2 column and run v3 alone. NOT TOUCHED: /z/scorecard_pipeline/ (the handoff said this path; it doesn't exist on Z \u2014 actual chiphazzard.py is at /z/setup_daily_reporting/reports/); pine v3 indicator source not pulled from trsports repo since the spec was complete enough to write v3 directly. Files added: /z/setup_daily_reporting/reports/chiphazzard_v3.py. Files modified: gen_mgp.py (import + _safe_chip_panel rewrite).",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "hot-key-managment-sell-of-postion",
      "phase": "design",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Hot Key Managment Sell % of Postion",
      "property": "alerts-hotkeys-etc",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-110118-6aa9ee",
      "note": "need to create hot keys to sell % of position, either it be 10% 25% 30% 50% 75% 80% 100% etc... .I'm thinking this can help with overtrading and being more objective on exits. And if mid-day trading can help stop churning with selling % of the position.",
      "project": "atlas"
    },
    {
      "id": "refactor-scientific-overlay-js-to-load-from-system",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Refactor _scientific_overlay.js to load from /system/strategies.json",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120000-7a3f1c",
      "note": "Currently /strategies/ page names _scientific_overlay.js as 'Source of truth'. Refactor it into a loader that fetches /system/strategies.json (schema v1) and dispatches by strategy id. No inline rule logic. First step of engine/UI separation per project_system_contract.md rule #5.",
      "project": "atlas"
    },
    {
      "id": "refactor-gen-scorecard-py-classify-entry-to-use-re",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Refactor gen_scorecard.py classify_entry() to use registry lookup",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120010-9b8d2e",
      "note": "classify_entry() has its own inline playbook logic that diverges from _scientific_overlay.js. Replace inline constants with lookups against /system/strategies.json so both batch (EOD scorecard) and live evaluator share one source.",
      "project": "multi_trader"
    },
    {
      "id": "split-td-vwap-from-vwap-pull-in-in-classifier",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Split TD VWAP from VWAP Pull-In in classifier",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120020-3e7a59",
      "note": "_scientific_overlay.js and gen_scorecard.py classify_entry() currently merge TD VWAP (pure Dale: flat bands, entry -1 SD) and VWAP Pull-In (James hybrid: entry VWAP to -0.5 SD). Schema v1 has them as separate registry entries (td-vwap-long, vwap-pull-in-long). Split the classifier so each gets distinct tagging; otherwise empirical buckets and live evals can't separate them.",
      "project": "multi_trader"
    },
    {
      "id": "add-strategy-id-column-to-indicator-alerts-main-jo",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "title": "Add strategy_id column to indicator_alerts + main.journal + backfill via classifier",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120030-c4f018",
      "note": "Phase 0 #5 enabler. Without strategy_id as a stable FK on trade/alert rows, downstream consumers (param surface, scorecard, live evaluator, accountability) cannot join cleanly. Backfill historical rows by running the new registry-driven classifier across them. Per system_contract.md SoT bullet #1.",
      "project": "multi_trader"
    },
    {
      "id": "refactor-strategies-page-to-render-from-system-str",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Refactor /strategies/ page to render from /system/strategies.json",
      "property": "strategy_library",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120040-5d291b",
      "note": "Today /strategies/ is hand-curated card list. Should be auto-rendered from the canonical registry: list all 5 + 2 entries, show entry rules, stop/target variants, empirical snapshot, citations. Filterable by family/direction. Bumps Strategy Library from a hand-maintained page (memory says 6d stale) to a generator (zero-stale).",
      "project": "strat_lib"
    },
    {
      "id": "build-drift-detector-diff-classifier-behavior-vs-r",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Build drift detector: diff classifier behavior vs registry vs Param Surface coverage",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120050-8f6c3a",
      "note": "Per system_contract.md non-negotiable rule #6: every generator validates strategy refs. The detector runs nightly, takes the registry as ground truth, and reports: classifier ids it emits that aren't in registry, registry strategies missing a Param Surface grid, registry strategies missing a classifier_function pointer, schema_version skew between artifacts. Alerts feed the freshness dashboard. [NIGHTLY] scan_strategy_drift.py (395 lines) + drift-report.{json,html} generated. Per system_contract.md rule #6 \u2014 every generator validates strategy refs.",
      "completed_at": "2026-05-12",
      "path": "/z/setup_daily_reporting/reports/scan_strategy_drift.py",
      "project": "atlas"
    },
    {
      "id": "promote-v1-1-schema-additions-author-live-eval-arr",
      "phase": "design",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Promote v1.1 schema additions: author live-eval arrays for all 7 strategies; bump schema_version to 2",
      "property": "strategy_library",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120100-2a9e4d",
      "note": "v1.1 draft (stress-tested against vwap-pull-in-long at /system/strategy-schema-v1_1-draft.md) adds validation_rules, invalidation_conditions, management_rules, quality_indicators per strategy. Author for the other 6 (td-vwap \u00d7 2, vwap-pull-in-short, trend-trade-long, backside \u00d7 2). Once all 7 are populated, bump meta.schema_version to 2 in strategies.json. [NIGHTLY] strategies.json bumped to schema_version 2 (confirmed by drift-report.json registry_schema_version=2, 9 strategies registered). v1.1 schema additions promoted.",
      "completed_at": "2026-05-12",
      "path": "/z/public_html/system/strategies.json",
      "project": "strat_lib"
    },
    {
      "id": "build-pure-function-evaluator-evaluate-py-live-bat",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "Build pure-function evaluator (evaluate.py) \u2014 live + batch grading function",
      "property": "accountability_hud",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120110-bc0357",
      "note": "Single function: evaluate(strategy_def, trade_state, market_context) \u2192 eval_record. No I/O, no side effects, testable with synthetic inputs. Loads /system/strategies.json once per run. Returns {grade, rules_satisfied, rules_failed, invalidation_triggered, quality_score, distance_from_optimal_exit, recommended_action, reasons[]}. Used by both the 60s live cron AND the EOD batch \u2014 that's how live grades stay coherent with post-hoc scorecard grades. Per system_contract.md non-negotiable rule #5. [NIGHTLY] evaluate.py landed (545 lines, pure-function). Same function consumed by 60s live cron and EOD batch per system_contract.md rule #5.",
      "completed_at": "2026-05-12",
      "path": "/z/setup_daily_reporting/reports/evaluate.py",
      "project": "acct_v2"
    },
    {
      "id": "ddl-cron-for-main-live-trade-evaluations-append-on",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "DDL + cron for main.live_trade_evaluations append-only eval log",
      "property": "accountability_hud",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120120-e418fa",
      "note": "Append-only table keyed by (trade_id, evaluated_at_ET, schema_version). Indexed on (trade_id, evaluated_at DESC) for latest-row reads. Cron writes a row per open trade every 60s during RTH; event-driven hooks (bar close, alert webhook, position-change) also append. Never overwrite \u2014 full audit trail of how a trade's grade evolved. [NIGHTLY] DDL written (112 lines) + live_eval_orchestrator.py (210 lines) wires the 60s append-only cron. Append-only audit log per spec.",
      "completed_at": "2026-05-12",
      "path": "/z/setup_daily_reporting/reports/live_trade_evaluations.sql",
      "project": "acct_v2"
    },
    {
      "id": "wire-live-grade-column-into-accountability-hud-v2-",
      "phase": "design",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Wire live grade column into Accountability HUD v2 open-positions section",
      "property": "accountability_hud",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120130-f7d62b",
      "note": "Per project_system_contract.md HUD-off-limits rule, this is a planned edit to live-accountability-v2 that needs James's explicit go-ahead before code changes. UI reads latest-row-per-trade from main.live_trade_evaluations and displays: live grade (A/B/C/F/OFF-BOOK), top failing rule, recommended action (hold/reduce/exit/trim). Prove the loop end-to-end on this one cockpit before fanning out to live-accountability-v2-dev.",
      "project": "acct_v2"
    },
    {
      "id": "fix-regime-classifier-0-trades-tagged-trend-in-467",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Fix regime classifier (0 trades tagged TREND in 467-entry sample)",
      "property": "scorecards",
      "created_at": "2026-05-12",
      "queued_from_form": true,
      "queued_id": "pending-20260512-120140-15e8c4",
      "note": "Per project_strategy_framework.md empirical block: 452 entries tagged ROTATION, 0 tagged TREND in the 2026-04-21 sample. TREND detection in classifier is broken. Hard blocker for live-evaluating trend-trade-long (its gating.regime_required is 'trend'; without a working classifier, the validation_rules will never satisfy and the strategy effectively can't be live-graded). Must land before evaluate.py is wired into prod for trend strategies. | Shipped 2026-05-13: regime_classifier_v2.py replaces v1 with slope_unit aware band-slope (sigma/bar, threshold 0.05) matching TraderDale_VWAP.pine. /system/classifier-divergence.html documents the divergence + fix.",
      "completed_at": "2026-05-13",
      "project": "atlas"
    },
    {
      "id": "live-eval-preview-page",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Live-eval preview page (read-only view of evaluate.py output, outside HUD)",
      "path": "/z/public_html/system/live-eval-preview.html",
      "completed_at": "2026-05-12",
      "created_at": "2026-05-12",
      "property": "accountability_hud",
      "note": "Standalone preview at /system/live-eval-preview.html to dogfood evaluate.py output before any HUD-v2 wire-up. Respects HUD-off-limits feedback rule by living under /system/.",
      "project": "acct_v2"
    },
    {
      "id": "mgp-recap-vital-briefings-rename",
      "phase": "ship",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "title": "MGP Recap email subject rebranded to 'Vital Briefings - End of Day Recap'",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_intraday.py",
      "completed_at": "2026-05-12",
      "created_at": "2026-05-12",
      "property": "mgp",
      "note": "Today's 16:37 recap shipped with subject 'Vital Briefings - End of Day Recap 2026-05-12' (was 'MGP Recap YYYY-MM-DD'). gen_mgp_intraday.py + mgp_email.py touched 20:52. Branding rename only; structure unchanged.",
      "project": "atlas"
    },
    {
      "id": "classify-v2-and-validation-rules-fixup",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "classify_v2.py + fix_validation_rules.py drafted (scaffolding for split + regime work)",
      "path": "/z/setup_daily_reporting/reports/classify_v2.py",
      "completed_at": "2026-05-12",
      "created_at": "2026-05-12",
      "property": "scorecards",
      "note": "classify_v2.py (290 lines) drafted as next-gen classifier; fix_validation_rules.py (98 lines) tightens schema-v2 validation. Does NOT yet complete the TD VWAP/Pull-In split or the TREND regime fix \u2014 those tasks remain pending. Captured as scaffolding so the work is visible on the progress calendar.",
      "project": "atlas"
    },
    {
      "id": "acct-v2-bug-ticket-meter-ignores-date-range",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "jimbo",
      "priority": "P0",
      "title": "BUG: Ticket Meter denominator doesn't scale with date range (shows 1901 / 30 on 13-day MTD range)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/trader-range.php",
      "created_at": "2026-05-13",
      "note": "Reported by James 2026-05-13 with screenshot from live DEV HUD. REPRO: select James + date range 2026-05-01 \u2192 2026-05-13 (MTD). OBSERVED: TICKET METER shows '1901 / 30' with the label '100% of daily ticket limit'. The numerator (1901) is correctly cumulative \u2014 sum of all tickets traded across the 13 days. The denominator (30) is the SINGLE-DAY max-trades limit and does NOT scale with the range. The '100%' display is also misleading because it caps at 100% even though 1901/30 = 6333% \u2014 actual overage is hidden by the bar's saturation. EXPECTED: when a multi-day range is selected, the denominator should sum the per-day limits across active trading days (e.g., 30 \u00d7 N active days = N\u00d730). For 3 ACTIVE DAYS visible elsewhere on the same screen, the denominator should be 90 \u2014 and 1901/90 = 2112% overage, which is the truth. SAME CLASS OF BUG as acct-v2-bug-risk-bump-ignores-date-range \u2014 both are 'denominator stays single-day-scoped, numerator goes cumulative' issues on the multi-day view. SUSPECTED CODE: api/trader-range.php (per-day limits not multiplied across range) OR hud/js/render.js (ticket meter renders the daily ticket_limit setting rather than the summed range limit). Also fix the 100% display ceiling so true overage is visible. SCREENSHOT: dir created at /todos/issues/acct-v2-bug-ticket-meter-ignores-date-range/ for evidence drop. CONTEXT FROM SAME SCREENSHOT: Daily Stop $130 declared 05/01 08:18 ET, Max Trades 30 self-imposed, Hard Stop $500 firm, Today's P&L +$92.18 across 1919 round trips, Net after commissions -$166.16. Best day in range +$80.28. Worst MDD -$145.50/-$130. Risk Bump bar ALSO showing the same scaling bug here: -$30.44/+$500 with '3 ACTIVE DAYS' \u2014 that's the previously-filed acct-v2-bug-risk-bump-ignores-date-range still open. CONSIDER: roll both these bugs into a single epic 'Date-range-aware HUD metrics' under jimbo if more denominator-scaling instances surface. ACCOUNTABILITY CODE OFF-LIMITS to Claude per feedback_hud_off_limits.md \u2014 Jim must approve and execute the fix. | CODE CONFIRMED 2026-05-14 (read-only investigation): bug is in hud/js/render.js lines 590-597. (1) Denominator: `tkCap = Number(t.max_trades || 30)` is the single-day cap, never multiplied by range length. FIX: render.js already has `t.trading_days_traded` (used line 984) and `t.risk_bump_active_days` (line 455) in the payload, so `tkCap = max_trades * trading_days_traded` needs no API change. (2) Display ceiling: `Math.min(100, ...)` at line 595 saturates the % text \u2014 remove the clamp from the text (bar-fill width can stay clamped). render.js was NOT edited in the 2026-05-13 HUD session, so the bug is still live.",
      "property": "accountability_hud",
      "project": "acct_v2",
      "sprint": 1
    },
    {
      "id": "daily-stop-intel-module-2026-05-13",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Daily Stop Intel module + briefing page (closes post-trade -> daily-stop loop)",
      "path": "/z/setup_daily_reporting/reports/daily_stop_intel.py",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "property": "daily-stop",
      "note": "daily_stop_intel.py reads personal_stats.json + positions_snapshot.json + strategies.json and emits /system/daily-stop-intel.json + /system/daily-stop-intel.html. Answers: which DML tier (tight/normal/relaxed) for tomorrow, why, which playbooks are costing.",
      "project": "multi_trader"
    },
    {
      "id": "mgp-independent-morning-pipeline-2026-05-13",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "MGP independent morning pipeline: catalysts + morning_tickers + morning_data (Vital Knowledge drop, phase 2)",
      "path": "/z/setup_daily_reporting/reports/morning_data.py",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "property": "mgp",
      "note": "Built drop-in replacement for vital_dawn_parser output. New: catalysts.py (yfinance earnings + news), morning_tickers.py (pre-market gapper scan against universe.json, no Vital Knowledge dependency), morning_data.py (assembles the dict shape gen_mgp.py expects). Updated: stocks_in_play.py, convergence.py, gen_mgp.py, gen_mgp_intraday.py, mgp_template.html. Verified by 4 successful MGP-triple regenerations between 17:34 and 19:44 ET. Parent task mgp-drop-vital-independent-source remains in_progress until James signs off on cutover.",
      "project": "atlas"
    },
    {
      "id": "system-strategy-system-page-2026-05-13",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "/system/strategy-system.html documentation page",
      "path": "/z/public_html/system/strategy-system.html",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "property": "system",
      "note": "New documentation surface under the system hub explaining the strategy registry / classifier / evaluator wiring.",
      "project": "atlas"
    },
    {
      "id": "issues-workspace-scaffold-2026-05-13",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "/todos/issues/ workspace scaffold (per-bug subdirectories)",
      "path": "/z/public_html/todos/issues/",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "property": "todo",
      "note": "Created /todos/issues/ as a workspace tree for accumulating per-bug context (repro notes, queries, screenshots). README + first subdir for acct-v2-bug-ticket-meter-ignores-date-range.",
      "project": "atlas"
    },
    {
      "id": "acct-v2-hud-dev-session-2026-05-13",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "jimbo",
      "title": "Accountability HUD v2 dev session \u2014 front-end pass + 2 API endpoints + KTI probe",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "property": "accountability_hud",
      "note": "File-scan-confirmed work session on the DEV HUD (James + Jim). Morning front-end pass: hud/js/main.js, hud/js/api.js, hud/js/drill.js, hud/index.html (11:32-11:35 ET). API endpoints: api/cto-clear.php, api/daily-settings.php (09:13). scripts/sync-realized-slow.js realized-P&L sync (11:03). Afternoon KTI investigation: scripts/kti-wait-probe.js (14:39) + kti-probe-run.ndjson / kti-probe-console.txt outputs (16:04). Prod tree untouched (cache JSON only) \u2014 consistent with read-only prod. Recorded from file evidence; specific deliverable scoping owned by Jim. NOTE: hud/js/render.js was NOT edited this session, so acct-v2-bug-ticket-meter-ignores-date-range + acct-v2-bug-risk-bump-ignores-date-range remain open.",
      "project": "acct_v2"
    },
    {
      "id": "etrade-positions-portfolio-script-2026-05-13",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "james",
      "title": "E*TRADE positions -> portfolio Google Apps Script",
      "path": "/z/trading/dev/a1p_journal/scripts/etrade-positions-to-portfolio-src.gs",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "note": "Google Apps Script (.gs) pulling E*TRADE positions into a portfolio sheet. Touched 13:18 ET. Recorded from file evidence; scope/intent owned by James.",
      "project": "atlas"
    },
    {
      "id": "traderdale-vwap-pine-update-2026-05-13",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "james",
      "title": "TraderDale_VWAP.pine canonical reference edit",
      "path": "/z/trading/indicators/src/indicators/TraderDale_VWAP.pine",
      "completed_at": "2026-05-13",
      "created_at": "2026-05-13",
      "note": "Canonical Dale VWAP Pine reference edited 13:44 ET, same afternoon as regime_classifier_v2.py (15:25). The classifier v2 aligns its band-slope formula (sigma/bar, threshold 0.05) to this file, so the two are connected work.",
      "project": "atlas"
    },
    {
      "id": "mgp-vwap-state-module",
      "phase": "build",
      "status": "completed",
      "title": "MGP: vwap_state.py building block for VWAP-state classification",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "New module replaces ad-hoc VWAP-state lookups in MGP generators; part of independent-pipeline phase.",
      "path": "/setup_daily_reporting/reports/vwap_state.py",
      "created_at": "2026-05-14",
      "property": "mgp",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "mgp-volume-zones-module",
      "phase": "build",
      "status": "completed",
      "title": "MGP: volume_zones.py + Pine companion (.pine) for volume-zone tagging",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "New module + Pine reference indicator. Pairs with stocks_in_play.py to surface zone-tagged setups in MGP.",
      "path": "/setup_daily_reporting/reports/volume_zones.py",
      "created_at": "2026-05-14",
      "property": "mgp",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "mgp-chiphazzard-v3",
      "phase": "build",
      "status": "completed",
      "title": "MGP: chiphazzard_v3.py replaces v1 (technical-bias upgrade)",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "v3 generator wired into gen_mgp.py / gen_mgp_intraday.py. Validates against breadth + sector-cycle inputs.",
      "path": "/setup_daily_reporting/reports/chiphazzard_v3.py",
      "created_at": "2026-05-14",
      "property": "mgp",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "playbook-amzn-2026-05-14",
      "phase": "ship",
      "status": "completed",
      "title": "Playbook: AMZN 2026-05-14 with chart-data extractor output",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "Generated via playbook pipeline (template + chart_data_extract.json). Card added to /playbooks/.",
      "path": "/public_html/playbooks/amzn-2026-05-14.html",
      "created_at": "2026-05-14",
      "property": "playbooks",
      "completed_at": "2026-05-14",
      "project": "multi_trader"
    },
    {
      "id": "playbook-nvda-2026-05-14",
      "phase": "ship",
      "status": "completed",
      "title": "Playbook: NVDA 2026-05-14 with charts.js companion",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "Same-day NVDA playbook. Card added to /playbooks/.",
      "path": "/public_html/playbooks/nvda-2026-05-14.html",
      "created_at": "2026-05-14",
      "property": "playbooks",
      "completed_at": "2026-05-14",
      "project": "multi_trader"
    },
    {
      "id": "playbook-qqq-rebrand",
      "phase": "ship",
      "status": "completed",
      "title": "Playbook: QQQ 2026-03-31 rebrand to house style",
      "owner": "James/Claude",
      "priority": "P3",
      "note": "Reference playbook rewritten; pre-debrand backup preserved on disk.",
      "path": "/public_html/playbooks/qqq-2026-03-31.html",
      "created_at": "2026-05-14",
      "property": "playbooks",
      "completed_at": "2026-05-14",
      "project": "multi_trader"
    },
    {
      "id": "playbook-best-april-2026",
      "phase": "ship",
      "status": "completed",
      "title": "Playbook: best-of-April-2026 retro selector output",
      "owner": "James/Claude",
      "priority": "P2",
      "note": "First real run of the best-trades selector (previously stubbed). Publishes /playbooks/best-april-2026/.",
      "path": "/public_html/playbooks/best-april-2026/",
      "created_at": "2026-05-14",
      "property": "playbooks",
      "completed_at": "2026-05-14",
      "project": "multi_trader"
    },
    {
      "id": "playbook-bella-fade-template",
      "phase": "scaffold",
      "status": "completed",
      "title": "Playbook: Bella Fade intern template scaffold",
      "owner": "James/Claude",
      "priority": "P3",
      "note": "Template directory for the SMB intern Bella Fade variant; ready for instances.",
      "path": "/public_html/playbooks/bella-fade-intern-template/",
      "created_at": "2026-05-14",
      "property": "playbooks",
      "completed_at": "2026-05-14",
      "project": "multi_trader"
    },
    {
      "id": "strategies-team-page",
      "phase": "ship",
      "status": "completed",
      "title": "Strategies: team-traders index page",
      "owner": "James/Claude",
      "priority": "P3",
      "note": "/strategies/team/ index added under the strategies hub.",
      "path": "/public_html/strategies/team/",
      "created_at": "2026-05-14",
      "property": "strategies",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "shared-tooltip-helper",
      "phase": "ship",
      "status": "completed",
      "title": "_shared/tooltip.js shared tooltip helper",
      "owner": "James/Claude",
      "priority": "P3",
      "note": "Shared tooltip module for cards / HUDs; replaces per-page inline tooltips.",
      "path": "/public_html/_shared/tooltip.js",
      "created_at": "2026-05-14",
      "property": "system",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "mgp-preview-staging",
      "phase": "ship",
      "status": "completed",
      "title": "MGP: /mgp/preview/ staging area for iterations",
      "owner": "James/Claude",
      "priority": "P3",
      "note": "Lets MGP regens be tested without overwriting the live morning page.",
      "path": "/public_html/mgp/preview/",
      "created_at": "2026-05-14",
      "property": "mgp",
      "completed_at": "2026-05-14",
      "project": "atlas"
    },
    {
      "id": "mgp-cron-skipped-5-14",
      "phase": "diagnose",
      "status": "pending",
      "title": "BUG: MGP auto-fire skipped 5/14 mid-day + recap; 5/15 morning ran ~2h45m late",
      "owner": "James",
      "priority": "P0",
      "note": "notify.log shows 5/14 06:41 send then nothing until 5/15 08:35 (normal slot 05:48). 5/14 mid-day + recap files exist but are manual evening regens that did not email. Two-day degradation. Inspect Task Scheduler + cron pipeline upstream of notify.log.",
      "path": "/setup_daily_reporting/reports/notify.log",
      "created_at": "2026-05-15T08:30:00-04:00",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "playbook-msft-2026-05-14",
      "phase": "phase-4",
      "status": "completed",
      "priority": "P1",
      "title": "Playbook: MSFT 2026-05-14 + charts.js companion",
      "owner": "James",
      "property": "playbooks",
      "path": "/playbooks/msft-2026-05-14.html",
      "completed_at": "2026-05-15",
      "note": "Published 17:06 ET alongside chart_data + msft-charts.js. Extends 5/14 playbook set from AMZN+NVDA to AMZN+NVDA+MSFT+XLE.",
      "project": "multi_trader"
    },
    {
      "id": "playbook-xle-2026-05-14",
      "phase": "phase-4",
      "status": "completed",
      "priority": "P1",
      "title": "Playbook: XLE 2026-05-14 + charts.js companion",
      "owner": "James",
      "property": "playbooks",
      "path": "/playbooks/xle-2026-05-14.html",
      "completed_at": "2026-05-15",
      "note": "Published 17:06 ET. Sector-ETF playbook complement to the single-name set.",
      "project": "multi_trader"
    },
    {
      "id": "strategies-family-migration",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "strategies.json family-field migration",
      "owner": "James",
      "property": "strategy_library",
      "path": "/system/strategies.json",
      "completed_at": "2026-05-15",
      "note": "Migration applied 16:17 ET. Pre-migration snapshot preserved at strategies.json.pre-family-migration.bak (78,419 -> 78,769 bytes). Aligns registry with canonical family taxonomy (VWAP / Pullback, Counter-trend, Level Breakout, Event-Driven, Gap Fade) referenced by UNIFIED-FRAMEWORK-v1.md.",
      "project": "strat_lib"
    },
    {
      "id": "unified-framework-v1-draft",
      "phase": "design",
      "status": "completed",
      "priority": "P1",
      "title": "UNIFIED-FRAMEWORK-v1.md doctrine draft",
      "owner": "James",
      "property": "foundation",
      "path": "/system/UNIFIED-FRAMEWORK-v1.md",
      "completed_at": "2026-05-15",
      "note": "1,165-line doctrine doc \u2014 composes the existing 13 Properties into a coherent strategy/scoring/optimization theory. Draft 2 (2026-05-15). Does NOT introduce new surfaces or redefine any property \u2014 pure architectural spec. Adjacent canonical docs: index.html (13-Property hub), strategy-schema-v1.md, strategy-schema-v1_1-draft.md, DECISIONS-streaming.md.",
      "project": "atlas"
    },
    {
      "id": "playbooks-modules-scanner-scaffold-2026-05-15",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "Playbooks: _modules/ + scanner/ subtree scaffold",
      "owner": "James",
      "property": "playbooks",
      "path": "/playbooks/",
      "completed_at": "2026-05-15",
      "note": "_modules/momentum_state_engine.{html,css} + scanner/{engine,index.html,scanner.json} scaffolded 16:58-17:05 ET. Sets up shared playbook tooling layer separate from per-symbol pages.",
      "project": "multi_trader"
    },
    {
      "id": "mgp-recap-5-15-tape-read-iterations",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "MGP code iterations: tape_read + mgp_template + gen_mgp + gen_mgp_intraday (2026-05-15 cycle)",
      "owner": "James",
      "property": "mgp",
      "path": "/setup_daily_reporting/reports/",
      "completed_at": "2026-05-15",
      "note": "Same-session edit cycle 12:07-13:38 ET: mgp_template.html, mgp_email.py, gen_mgp.py, _tape_smoke.py, tape_read.py, gen_mgp_intraday.py. Drove the 14:14 ET recap (chop-day tagged) which ran ahead of the normal 16:37 slot. Recap regen at 16:38 ET. Pre-cutover hardening for the independent MGP pipeline.",
      "project": "atlas"
    },
    {
      "id": "mgp-5-14-recap-backfill",
      "phase": "phase-mgp",
      "status": "completed",
      "priority": "P1",
      "title": "Backfill missing 5/14 MGP recap page (silent failure recovery)",
      "owner": "James",
      "property": "mgp",
      "path": "/mgp/2026-05-14-recap.html",
      "completed_at": "2026-05-15",
      "note": "Generated 13:27 ET on 5/15 to fill the 5/14 mid-day+recap gap (notify.log silent failure). Page recovered; root cause of the cron skip still under diagnose (see mgp-cron-skipped-5-14).",
      "project": "atlas"
    },
    {
      "id": "monthly-may-2026-publish",
      "phase": "phase-mgp",
      "status": "completed",
      "priority": "P1",
      "title": "May-2026 monthly report + VWAP-bands param surface full regen",
      "owner": "James",
      "property": "monthly_reports",
      "path": "https://smb.a1prospect.com/db_reports/james-monthly-reports/may-2026/",
      "completed_at": "2026-05-15",
      "note": "Full pipeline regen 16:30-16:46 ET: may-2026/index.html, vwap-bands/index.html, vwap-bands/param-surface/ (15 symbol surfaces+heatmaps, 7 TOD windows, tod_comparison, tod_sd_expansion), james-monthly-reports/index.html, vwap-research/index.html. Auto-pipeline output; ran clean.",
      "project": "atlas"
    },
    {
      "id": "ams-integration-plan-v1-draft",
      "phase": "design",
      "status": "completed",
      "priority": "P0",
      "title": "AMS (Accountability Management System) Integration Plan v1 draft",
      "owner": "James",
      "property": "foundation",
      "path": "/system/AMS-INTEGRATION-PLAN-v1.md",
      "completed_at": "2026-05-16",
      "note": "42KB integration doctrine \u2014 sits alongside UNIFIED-FRAMEWORK-v1.md. UF = what the platform IS; AMS plan = HOW existing pieces become one cohesive product. 17-week phased plan (5/18 -> 9/11) unifying 11 surfaces + porting auth from a1p_journal. Phase 0 bootstrap done. 7 open decisions in plan section 8.",
      "project": "atlas"
    },
    {
      "id": "todo-audit-v6-2026-05-16",
      "phase": "design",
      "status": "completed",
      "priority": "P1",
      "title": "Todo Audit v6: 9-pass classification + dual weekend pick queues",
      "owner": "claude",
      "property": "pm_system",
      "path": "/todos/audit/2026-05-16/",
      "completed_at": "2026-05-16",
      "note": "110 pending tasks classified across 9 passes: classification, dependencies, compounding infra, reversibility, done-when/horizon/plain-lang, drift (12 findings, 5 trace to catalog ownership), entropy, cognitive debt. Dual queues: autonomous Saturday (9 picks) + supervised (2 + 2 conditional). HEADLINE: catalog-refactor-shared-json (Tier 1 compounding) for Sunday pair work. Consistency-check pass caught 3 named inconsistencies + 1 cascade \u2014 all resolved. Bundle: 136KB.",
      "project": "atlas"
    },
    {
      "id": "platform-integration-plan-2026-audit-companion",
      "phase": "design",
      "status": "completed",
      "priority": "P1",
      "title": "Platform Integration Plan (2026) \u2014 audit companion doc",
      "owner": "claude",
      "property": "foundation",
      "path": "/todos/audit/2026-05-16/platform-integration-plan-2026.md",
      "completed_at": "2026-05-16",
      "note": "16KB phased roadmap (tables + Mermaid diagrams) under todos/audit. Sits adjacent to the larger /system/AMS-INTEGRATION-PLAN-v1.md as the audit-derived view of the same integration.",
      "project": "atlas"
    },
    {
      "id": "strategies-v1_2-schema-draft",
      "phase": "design",
      "status": "completed",
      "priority": "P1",
      "title": "strategies.json v1.2 schema stress-test draft",
      "owner": "claude",
      "property": "strategy_library",
      "path": "/system/strategies-v1_2-draft.json",
      "completed_at": "2026-05-16",
      "note": "22KB stress-test draft of next strategies-registry schema. Iterates on v2 (post-family-migration) \u2014 explores added fields needed to close the page-vs-registry coverage gap (19 page setups vs 8 registry strategies). Companion: PAGE-VS-JSON-AUDIT.md.",
      "project": "strat_lib"
    },
    {
      "id": "strategies-page-vs-json-audit",
      "phase": "design",
      "status": "completed",
      "priority": "P1",
      "title": "Page vs JSON coverage audit (/strategies/ vs strategies.json)",
      "owner": "claude",
      "property": "strategy_library",
      "path": "/system/PAGE-VS-JSON-AUDIT.md",
      "completed_at": "2026-05-16",
      "note": "Documents the gap: the /strategies/ page (and _scientific_overlay.js mirror) carries 19 setups; the evaluator registry knows 8 (+1 off-book sentinel). Trader can pattern-match against 19; evaluator can grade 8. Every page-only setup is silently OFF-BOOK. Names every divergent strategy and identifies trend-trade-short as the lone registry-only entry. Drives v1.2 schema work.",
      "project": "strat_lib"
    },
    {
      "id": "strategy-drift-scanner",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "Strategy drift scanner: scan_strategy_drift.py + drift-report.json/html",
      "owner": "claude",
      "property": "strategy_library",
      "path": "/system/drift-report.html",
      "completed_at": "2026-05-16",
      "note": "Automated scanner that walks strategies.json registry vs every mirror surface (_scientific_overlay.js etc.) and reports missing/extra ids. Output: drift-report.json + drift-report.html (regenerated 11:13 ET). Status today: 7 registry ids missing from _scientific_overlay.js mirror \u2014 confirms PAGE-VS-JSON-AUDIT findings programmatically.",
      "project": "strat_lib"
    },
    {
      "id": "trade-audit-replay-v1-vs-v2",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "Trade audit replay: classify v1 (band-blind JS) vs classify_v2.py (Pine/sigma)",
      "owner": "claude",
      "property": "strategy_library",
      "path": "/system/trade-audit-replay.html",
      "completed_at": "2026-05-16",
      "note": "audit_trades.py (regenerated 11:16 ET) replays the 5 synthetic-example trades through both classifiers side-by-side. Today's run: 1/5 regime divergence, 0/5 playbook divergence, 1 off-book under v2. Surfaces the v1/v2 classifier split documented in project_strategy_classifier_divergence.md.",
      "project": "strat_lib"
    },
    {
      "id": "strategy-system-landing-page-iteration",
      "phase": "build",
      "status": "completed",
      "priority": "P2",
      "title": "/system/strategy-system.html iteration",
      "owner": "claude",
      "property": "strategy_library",
      "path": "/system/strategy-system.html",
      "completed_at": "2026-05-16",
      "note": "Landing page for the strategy-classification subsystem (registry + drift report + audit replay). Touched 11:18 ET as part of today's strategy-tooling pass.",
      "project": "strat_lib"
    },
    {
      "id": "may-2026-monthly-full-regen-5-16",
      "phase": "build",
      "status": "completed",
      "priority": "P1",
      "title": "May-2026 monthly + Param Surface full regeneration",
      "owner": "claude",
      "property": "monthly_reports",
      "path": "https://smb.a1prospect.com/db_reports/james-monthly-reports/",
      "completed_at": "2026-05-16",
      "note": "Full pipeline regen 16:30-16:32 ET: monthly index, vwap-bands index, param-surface index, all 14 per-symbol heatmap+surface PNGs (MSFT/SNDK/USO/BE/CAR/ORCL/NVDA/WULF/GOOGL/ETHU/DUOL/IBIT/INTC/TSLA/PLTR), all 7 TOD heatmap+surface PNGs (open/early/mid_am/midday/mid_pm/late + comparison + sd_expansion). Saturday cadence \u2014 fresh material for Sunday review.",
      "project": "atlas"
    },
    {
      "id": "ams-phase1-auth-surface",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Phase 1: auth surface scaffolded (10 endpoints under /ams/_auth)",
      "path": "/z/gt_dev/ams/_auth/",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "Login HTML + session_guard + admin_guard + email_sender + check_email + check_username + accept_invite + forgot_password + reset_password landed in /z/gt_dev/ams/_auth/. Mirrors to /ams/_auth/ in public_html. Phase 1 of the 17-week integration plan, kicked off a day early (plan officially starts Monday 5/18).",
      "project": "atlas"
    },
    {
      "id": "ams-phase1-api-surface",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Phase 1: _api surface (entitlements, notifications, scope)",
      "path": "/z/gt_dev/ams/_api/",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "entitlements.php, notifications.php, scope.php landed in /z/gt_dev/ams/_api/. Reads ams.user_service_access registry \u2014 first cut of the service-entitlement contract called out in AMS Integration Plan v1 \u00a73.",
      "project": "atlas"
    },
    {
      "id": "ams-phase1-app-surface",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Phase 1: app surface (launcher + service detail + settings + trial flow)",
      "path": "/z/gt_dev/ams/app/",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "app/index.php (single-zone launcher), service.php, start_trial.php, settings.php, settings_save.php, notifications.php. Carries the per-user service grant + trial-grant flow. Two-zone launcher + per-service detail pages still pending (called out as Stage 2 in the new commercial roadmap).",
      "project": "atlas"
    },
    {
      "id": "ams-phase1-shell-header",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS Phase 1: shared shell header (_shell/header.php)",
      "path": "/z/gt_dev/ams/_shell/header.php",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "First piece of the cohesive AMS shell. Topnav + user menu + banner zone. Will wrap legacy surfaces (HUD, MGP, Scorecards) in Stage 2.",
      "project": "atlas"
    },
    {
      "id": "ams-commercial-roadmap-v1",
      "phase": "phase-ams-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Commercial Roadmap v1 \u2014 5-stage evolution + SaaS vision",
      "path": "/z/public_html/system/AMS-COMMERCIAL-ROADMAP.md",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "Captures POC -> internal hardened -> per-service paid -> multi-broker -> orgs/B2B stages. Documents what's already built (~80% of vision supported without rewrites), the gap to vision (cohesive shell, two-zone launcher, trial lifecycle, license UI, orgs, Stripe), and what to decide NOW vs LATER. Living doc.",
      "project": "atlas"
    },
    {
      "id": "ams-architecture-diagrams-v1",
      "phase": "phase-ams-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS Architecture Diagrams \u2014 system / DB-ER / user-flow (3 views)",
      "path": "/z/public_html/system/AMS-ARCHITECTURE-DIAGRAMS.md",
      "property": "ams",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "Mermaid diagrams: (1) system layers + external integrations, (2) database ER across main/ams/accountability schemas, (3) invite-issue -> service-access critical user-flow. Companion to Integration Plan v1 and Commercial Roadmap.",
      "project": "atlas"
    },
    {
      "id": "may-2026-monthly-regen-5-17",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "title": "May 2026 monthly report + VWAP param surface full regen (16 symbols, 6 TOD slots)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/",
      "property": "monthly_reports",
      "completed_at": "2026-05-17",
      "created_at": "2026-05-17",
      "note": "May 2026 index + VWAP bands index + Param Surface DB index regenerated 16:30 ET. 16 per-symbol heatmaps + surfaces (MSFT, SNDK, USO, BE, CAR, ORCL, NVDA, WULF, GOOGL, ETHU, DUOL, IBIT, INTC, TSLA, PLTR, plus combined). 6 TOD slots (Open, Early, Mid-AM, Midday, Mid-PM, Late) + TOD comparison + SD expansion. Auto-pipeline output, no source code changes.",
      "project": "atlas"
    },
    {
      "id": "ams-cockpit-port",
      "phase": "phase-ams-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "AMS Phase 2: full Cockpit/HUD code port into /ams/cockpit/ (CSS + 15 JS modules + 27 API endpoints + SSE)",
      "path": "/z/gt_dev/ams/cockpit/",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "Massive port: hud.css, 15 JS modules (apply-env, breaches-modal, calendar-modal, daily-settings, drill, events, main, scorecard-click, scorecard-grade, scorecard-chart, strategies-fuzzy, table-sort, api, render, usage-report), env config (env.dev.json/env.prod.json/env.json with README), 27 _api/cockpit/ endpoints (ingest, kti-live, php-config-check, refresh-trader-settings, trigger-backfill, _guard, traders, trader, trader-range, trades, scorecard-trade, scorecard-leg, calendar, breaches, chart-bars, bridge-symbols, data-coverage, health, cto-clear, daily-settings, trade-action, stream/SSE, usage-event, usage-report, bootstrap), favicon, index.php, usage-report.html. This is Stage 2 cohesive-shell wrap in flight \u2014 supersedes yesterday's ams-shell-stage2-decision task_id by execution. Next: end-to-end smoke + cutover plan (see tomorrow's TODAY #1).",
      "project": "multi_trader"
    },
    {
      "id": "ams-admin-users-teams-licenses",
      "phase": "phase-ams-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS admin panel scaffolded: users + teams + licenses",
      "path": "/z/gt_dev/ams/admin/",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "admin/users/ (index, edit, new_from_trader, issue_reset, set_temp_password), admin/teams/ (edit), admin/licenses/ (placeholders + error_log). First admin surface for the multi-user platform \u2014 enables operator workflows around trader onboarding, password resets, and team management.",
      "project": "atlas"
    },
    {
      "id": "ams-gchat-webhook",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS Groups gchat webhook: SQL migration 012 + _api/gchat.php",
      "path": "/z/gt_dev/ams/_api/gchat.php",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "sql/migrations/012_groups_gchat_webhook.sql adds gchat_webhook column; _api/gchat.php endpoint posts to it. Notification path for team Google Chat alerts (Domino Brief, breaches, etc.).",
      "project": "atlas"
    },
    {
      "id": "ams-daily-settings-save",
      "phase": "phase-ams-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS daily-settings save path: _api endpoint + cockpit JS module",
      "path": "/z/gt_dev/ams/_api/daily_settings_save.php",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "_api/daily_settings_save.php + _api/cockpit/daily-settings.php + cockpit/js/daily-settings.js. Per-trader per-day setting persistence (loss limits, sizing overrides, regime overrides) wired through the AMS API layer instead of the legacy accountability/config.json path.",
      "project": "atlas"
    },
    {
      "id": "ams-auth-login-php",
      "phase": "phase-ams-1",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS auth: login.html \u2192 login.php upgrade (server-side handler) + whoami.php",
      "path": "/z/gt_dev/ams/_auth/login.php",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "Yesterday's static login.html now backed by login.php server handler (form POST \u2192 session); whoami.php returns the authenticated user for JS layer. Closes the auth round-trip that was scaffolded but not wired yesterday.",
      "project": "atlas"
    },
    {
      "id": "ams-legacy-mounts",
      "phase": "phase-ams-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS legacy-surface mounts: mgp/playbooks/scorecards/system into /ams/",
      "path": "/z/gt_dev/ams/",
      "property": "ams",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "mgp/, playbooks/, scorecards/, system/ added as symlinks/junctions into the AMS tree at 23:16 ET. Foundation for routing legacy surfaces through the AMS shell (Stage 2 cohesive-shell wrap). Scanner can't recurse through these (reparse points) \u2014 pages still live at their canonical /mgp /playbooks /scorecards /system URLs.",
      "project": "atlas"
    },
    {
      "id": "playbooks-bb-rsi-rollout",
      "phase": "phase-parallel",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Playbooks: Bollinger Bands + RSI sub-chart pattern rolled out (AMZN, MSFT, NVDA, XLE)",
      "path": "/z/public_html/playbooks/",
      "property": "playbooks",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "AMZN, MSFT, NVDA, XLE chart pages updated with BB + RSI sub-chart (matches reference_chart_defaults.md). Each has a .bak-pre-bb-rsi-2026-05-18 snapshot for rollback. Also: new SOXL 2026-05-18 playbook + soxl-charts.js + soxl_chart_data.json; MSFT 2026-05-14 playbook regenerated with new chart_data.json; new _modules/momentum_state_engine.html shared component.",
      "project": "multi_trader"
    },
    {
      "id": "vwap-research-index",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "title": "New vwap-research/index.html tree under gt_dev/db_reports",
      "path": "/z/gt_dev/db_reports/vwap-research/",
      "property": "param_surface",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "Created at 16:46 ET (right after the 16:30 monthly regen). Standalone landing for VWAP-band research work \u2014 separate from the param-surface DB output. First seed of what may become the strategy-research surface.",
      "project": "strat_lib"
    },
    {
      "id": "may-2026-monthly-regen-5-18",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "title": "May 2026 monthly + VWAP Param Surface full regen (16 symbols, 6 TOD slots) \u2014 Monday cadence",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/",
      "property": "monthly_reports",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "16:30 ET auto-pipeline run. Same 16-symbol + 6-TOD-slot output as 5/17. No source changes.",
      "project": "atlas"
    },
    {
      "id": "mgp-cycle-5-18",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "title": "MGP all 3 editions on schedule \u2014 Morning 'lean-risk-off \u00b7 breadth -2', Mid-Day 'thesis broken', Recap 'reversal day'",
      "path": "/z/public_html/mgp/",
      "property": "mgp",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "Morning 05:41 ET 'lean-risk-off \u00b7 breadth -2 \u00b7 yellow lights'. Mid-Day 11:38 'thesis broken'. Recap 16:37 'reversal day'. The thesis-broken\u2192reversal narrative tracks Friday's chop-day setup falling through. notify.log clean (no timeouts, no retries).",
      "project": "atlas"
    },
    {
      "id": "hud-config-edit-5-18",
      "phase": "phase-hold",
      "status": "completed",
      "owner": "user",
      "priority": "P3",
      "title": "James edited accountability/config.json via cockpit UI (2 backups, 09:30 ET final)",
      "path": "/z/gt_dev/db_reports/smb-reports-v2/accountability/config.json",
      "property": "accountability_hud",
      "completed_at": "2026-05-18",
      "created_at": "2026-05-18",
      "note": "Two edits via the cockpit UI: 09:19 ET (config-before-edit-2026-05-18T13-19-47Z.json snapshot) and 09:30 ET (final state, config-before-edit-2026-05-18T13-30-45Z.json). User-driven, off-limits rule respected. Worth flagging because the new ams-daily-settings-save path will eventually supersede this file as the persistence layer for per-day settings.",
      "project": "acct_v2"
    },
    {
      "id": "ams-phase2-data-scope-privacy",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS Phase 2: data scope + privacy (migrations 010+011, scope.php, settings + admin link)",
      "owner": "Claude",
      "priority": "P0",
      "property": "system_contract",
      "note": "Migration 010 (main.users.preferences JSONB) + 011 (user_view_scope/user_trade_scope SQL fns + can_view_* gates). _api/scope.php helpers (ams_view_scope_user_ids, ams_trade_scope_account_ids). app/settings.php Trading-profile card lets user self-edit willing_to_share + short_name. admin/users/edit.php has search-as-you-type trader picker + admin share override. Verified: user_trade_scope(6)=[14859,14959]; user_trade_scope(18)=all 6 (super_admin).",
      "project": "atlas"
    },
    {
      "id": "ams-phase2_5-gchat-fanout",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS Phase 2.5: Google Chat fan-out (migration 013 promotes webhook to teams + override path)",
      "owner": "Claude",
      "priority": "P1",
      "property": "system_contract",
      "note": "Migration 013 promotes gchat_webhook_url from groups \u2192 teams (per-team default + per-group override via COALESCE). _api/daily_settings_save.php persists + best-effort fan-outs via ams_gchat_post_for_user(). admin/teams/edit.php has team-webhook field with per-group override framing. Pending: James pastes the actual Home-team incoming-webhook URL + does end-to-end verify.",
      "project": "atlas"
    },
    {
      "id": "ams-phase3a-cockpit-rebuild",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS Phase 3a: cockpit REBUILD at /ams/cockpit/ \u2014 v2 port + scope/entitlement gates on every endpoint",
      "owner": "Claude",
      "priority": "P0",
      "property": "system_contract",
      "note": "Full v2 source mirror at /ams/cockpit/ + 22 endpoints at /ams/_api/cockpit/. Shared scope guard (_guard.php) provides cockpit_require_session/cockpit_require_short/cockpit_account_in_clause. Grid filter via accountability.v_trader_today WHERE account_id IN scope. Every read endpoint (trader, trader-range, calendar, trades, breaches, chart-bars, scorecard-*, data-coverage, bridge-symbols, health) and every write endpoint (cto-clear, daily-settings, trade-action) cockpit_require_short()-gated \u2192 403 out-of-scope. SSE stream snapshots scope at connect + drops account-level NOTIFYs not in scope. Launcher service URL \u2192 /ams/cockpit/; missing-entitlement banner offers trial-start. Verified end-to-end as super_admin jimbo (full grid) vs scoped user grumpy (2 rows, 403 on out-of-scope query).",
      "project": "multi_trader"
    },
    {
      "id": "ams-cockpit-cutover-smoke",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS Cockpit cutover smoke: head-to-head /ams/cockpit/ vs live-accountability-v2",
      "owner": "Claude",
      "priority": "P0",
      "property": "system_contract",
      "note": "Listed as TODAY priority #1 in yesterday's nightly report. Closed by Phase 3a rebuild: end-to-end smoke documented in PHASE-3a-STATUS.md \u00a7Tangible verification. Cutover is non-destructive \u2014 launcher now points at /ams/cockpit/, but /live-accountability-v2/ stays mounted untouched (HUD-off-limits rule honored, ADR-0001 honored). No traffic redirect/rewrite yet \u2014 users land on /ams/cockpit/ via the launcher tile.",
      "project": "atlas"
    },
    {
      "id": "ams-cockpit-calendar-net-pnl",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "Cockpit calendar: commissions + net P&L tiles + per-day cards (was gross-only)",
      "owner": "Claude",
      "priority": "P2",
      "property": "system_contract",
      "note": "cockpit/js/render.js extended \u2014 summary tiles + per-day cards now display commission + net alongside gross. Lands in /ams/cockpit/ port; not back-ported to legacy v2 tree (HUD-off-limits rule + parallel-build pattern).",
      "project": "atlas"
    },
    {
      "id": "ams-auth-whoami-redirect",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS auth: whoami.php endpoint + already-signed-in redirect from login page",
      "owner": "Claude",
      "priority": "P1",
      "property": "system_contract",
      "note": "_auth/whoami.php returns {logged_in,user_id,role,preferred_landing_url}. login.html now calls whoami() on load \u2014 if session is valid, bounces straight to the user's preferred landing (cockpit/launcher) \u2014 no more double-login. Login default landing flipped from launcher \u2192 cockpit (entitlement-aware fallback when no grant).",
      "project": "atlas"
    },
    {
      "id": "ams-session-heartbeat-idle-mode",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS session heartbeat + strict idle mode + cross-tab config sync",
      "owner": "Claude",
      "priority": "P2",
      "property": "system_contract",
      "note": "_shell/header.php now polls /ams/_auth/check.php every 60s; 401 \u2192 bounce to login with ?expired=1&return=. Strict idle mode (user preference) tracks mouse/keyboard/scroll + forces logout after configured timeout regardless of background XHR. Cross-tab config sync via localStorage broadcast/listener so settings changes propagate to all open AMS tabs. Bonus: header $shell_user collision fixed \u2014 topnav pill always uses $_SESSION (was misrendering target user on admin edit pages).",
      "project": "atlas"
    },
    {
      "id": "ams-profile-full-cards",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS user profile: 5 cards (Account/Address/Public/Professional/Enterprise) + admin mirror + admin ops",
      "owner": "Claude",
      "priority": "P1",
      "property": "system_contract",
      "note": "settings.php + app/settings_save.php: Account & contact (username/email/phone \u2014 email change forces re-verify), Address, Public profile, Professional profile, Enterprise profile. admin/users/edit.php mirrors all 5 + Admin Ops card (unlock account / force email reverify / reset 2FA). Plus create-user-from-trader flow: banner on user-list surfaces orphan traders + one-click prefilled new_from_trader.php with auto-link + reset-link handoff. Multi-select add-members on admin/teams/edit.php (checkbox grid + select-all-visible + search filter + count-aware submit). Search/filter on admin user list + search-as-you-type trader picker.",
      "project": "atlas"
    },
    {
      "id": "ams-email-deliverability-dmarc-diag",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "Email deliverability: DMARC added (monitor mode) + super_admin diag_send.php tool",
      "owner": "Claude",
      "priority": "P0",
      "property": "system_contract",
      "note": "DMARC _dmarc.a1prospect.com added 2026-05-18: p=none; rua=mailto:postmaster@... ; aspf=r; adkim=r; pct=100 (monitor mode, per MGP-deliverability memory rule p=none). _auth/diag_send.php (super_admin only) sends to mail-tester.com + displays current SPF/DKIM/DMARC state. STILL PENDING (manual user steps): A2 cPanel DKIM Repair, create postmaster@ forwarder, run diag + paste mail-tester score. Until those 3 land, invite/reset emails likely still go to spam \u2014 Phase 1 first-trial flow is blocked on it.",
      "project": "atlas"
    },
    {
      "id": "ams-launcher-cockpit-url-bind",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS launcher: cockpit tile URL bound to /ams/cockpit/ + missing-service banner",
      "owner": "Claude",
      "priority": "P1",
      "property": "system_contract",
      "note": "UPDATE ams.services SET url='/ams/cockpit/' WHERE service_id='accountability_hud'. Launcher (app/index.php) shows yellow banner with one-click trial-start when entitlement gate redirects user back.",
      "project": "atlas"
    },
    {
      "id": "ams-mgp-fork-prep",
      "phase": "design",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "MGP-AMS fork prep: gen_mgp_ams.py + gen_mgp_intraday_ams.py + notify_ams.py + recipients",
      "owner": "Claude",
      "priority": "P1",
      "property": "mgp",
      "note": "Forked generators landed in /z/setup_daily_reporting/reports/ at ~19:40 ET \u2014 gen_mgp_ams.py / gen_mgp_intraday_ams.py / notify_ams.py / mgp_recipients_ams.json. Foundation for routing MGP through AMS entitlement + per-user scope rather than the hardcoded jdomino/Jim.Domino recipient list. Not yet wired to cron \u2014 live MGP today still went through the legacy gen_mgp.py path (Morning 05:40, Mid-Day 11:38, Recap 16:37 all delivered).",
      "project": "atlas"
    },
    {
      "id": "ams-stage-tree-populated",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "AMS stage tree /trading/ams/ populated for the first time (canonical stage path per AMS-paths memory)",
      "owner": "Claude",
      "priority": "P1",
      "property": "system_contract",
      "note": "Mirror of dev AMS tree landed in /z/trading/ams/ at 11:32 ET \u2014 full _api/cockpit/ surface, _auth/, _shell/, admin/, app/, cockpit/, mgp/playbooks/scorecards/system symlinks. Per AMS-paths memory: /trading/ams/ IS the canonical stage path (vs /trading/dev/ams/ = canonical dev). Stage now has a real baseline against which dev \u2192 stage promotion can be diffed.",
      "project": "atlas"
    },
    {
      "id": "vwap-research-index-new",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-05-19",
      "title": "New /vwap-research/ index page on db_reports",
      "owner": "Claude",
      "priority": "P2",
      "property": "monthly_reports",
      "note": "New index.html landed at gt_dev/db_reports/vwap-research/ at 16:46 ET. Scope/purpose TBD \u2014 surface for VWAP-band research that doesn't fit cleanly inside the Param Surface property.",
      "project": "atlas"
    },
    {
      "id": "alpha-trend-fix-mgp-manual-recovery-2026-05-21",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "alpha_trend.py fix + venv rebuild + manual MGP regen (5/21 morning pipeline failed)",
      "path": "/z/setup_daily_reporting/reports/alpha_trend.py",
      "completed_at": "2026-05-21",
      "created_at": "2026-05-21",
      "note": "5/21 nightly/morning MGP cron did not deliver (notify.log has ZERO 5/21 entries). In the late afternoon alpha_trend.py was edited (17:17 ET) and the reports .venv rebuilt (17:18), then all three 5/21 MGP pages were regenerated by hand (morning 17:23, midday 17:33, recap 17:50) \u2014 HTML now exists at /mgp/2026-05-21*.html but the emails were never sent. Likely fallout from the in-flight Vital->independent-source migration (see mgp-drop-vital-independent-source). Recovery restored the published pages; email delivery for 5/21 remains a gap.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "accountability-config-edits-cockpit-2026-05-22",
      "phase": "phase-3",
      "status": "completed",
      "owner": "james",
      "title": "Accountability config edits via cockpit admin UI (5/20 + 5/22, with backups)",
      "path": "/z/gt_dev/db_reports/smb-reports-v2/accountability/config.json",
      "completed_at": "2026-05-22",
      "created_at": "2026-05-20",
      "note": "config.json edited through the AMS-side cockpit admin path on 5/20 09:31 ET and 5/22 08:51 ET, each with a timestamped backup written to accountability/backups/. HUD-off-limits rule respected (edits via admin UI, not source). Bookkeeping only.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "nightly-pm-silent-gap-2026-05-20-to-22",
      "phase": "phase-0",
      "status": "pending",
      "owner": "james",
      "priority": "P0",
      "title": "Nightly PM daemon went dark 5/20-5/22 (3 consecutive missed fires) \u2014 fix wake/launch",
      "path": "/z/logs/nightly-startup.log",
      "created_at": "2026-05-23",
      "note": "The accountability system's own nightly reporter did not run on 5/20 (Wed), 5/21 (Thu), or 5/22 (Fri). nightly-startup.log jumps straight from 5/19 23:21 DONE to 5/23 23:05 LAUNCHED \u2014 no LAUNCHED line for the three intervening nights. This is the documented machine-sleep / scheduled-tasks-daemon-launch failure mode (reference_automation_runtime): WTS task needs jtdom logged in + Claude Code open. Three weeknights in a row is worse than the usual weekend slip. Backfilled tonight (5/20-5/22 reports written), but the root cause (auto-login / wake-from-sleep) is still James's open call. | UPDATE 2026-05-25: 5/24 (Sun) ALSO silently missed \u2014 nightly-startup.log jumps 5/23 DONE -> 5/25 LAUNCHED with no 5/24 LAUNCHED line. Now 5 silent misses in ~2 weeks (5/20, 5/21, 5/22, 5/24). Still backfilled-only; root cause unchanged.",
      "property": "todos",
      "project": "atlas"
    },
    {
      "id": "mgp-delivery-regression-2026-05-21-22",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "title": "MGP delivery regression: 5/21 sent zero emails, 5/22 recap never generated",
      "path": "/z/setup_daily_reporting/reports/notify.log",
      "created_at": "2026-05-23",
      "note": "Two MGP delivery failures in the backfill window. (1) 5/21: all three HTML pages exist but notify.log has NO 5/21 entries at all \u2014 nothing emailed; pages were hand-regenerated at 17:23-17:50 after an alpha_trend.py fix. (2) 5/22: morning fired ~2h late (07:49 vs usual ~05:41) and the End-of-Day Recap never generated (no /mgp/2026-05-22-recap.html, no notify entry) \u2014 only morning + midday delivered. Both line up with the in-flight Vital->independent-source migration destabilizing the cron path. Stabilize delivery (and the recap cron) BEFORE switching cron fully to --source=independent or cancelling Vital.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "ams-scorecards-rebuild",
      "phase": "ship",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Scorecards property rebuilt at /ams/scorecards/ (scope-guarded port, dev + stage)",
      "path": "/z/trading/dev/ams/scorecards/",
      "completed_at": "2026-05-20",
      "created_at": "2026-05-20",
      "note": "Was 5/19's TODAY #2. New scorecards/api/live.php + scorecards/live.php + scorecards/index.php built in dev 03:49-06:00 ET, then synced to stage /trading/ams/scorecards/ 05:29-06:01. Second property (after Cockpit) ported into the scope-guarded AMS shell. Recovered via supplemental /trading/dev/ams scan \u2014 SKILL's default scan roots don't cover the AMS tree.",
      "property": "scorecards",
      "project": "multi_trader"
    },
    {
      "id": "ams-cockpit-scorecard-drilldown",
      "phase": "ship",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS Cockpit scorecard drill-down + chart-bars endpoint (dev + stage)",
      "path": "/z/trading/dev/ams/cockpit/js/drill.js",
      "completed_at": "2026-05-21",
      "created_at": "2026-05-21",
      "note": "5/21 14:42-16:16 ET. New _api/cockpit/trader.php + _api/cockpit/chart-bars.php; cockpit JS drill.js + scorecard-chart.js + scorecard-click.js; cockpit/css/hud.css; cockpit/index.php. Built in dev, synced to stage /trading/ams/ 16:08-16:16. Clicking a scorecard cell in the AMS cockpit now opens the drill chart. Recovered via supplemental /trading/dev/ams scan.",
      "property": "accountability_hud",
      "project": "acct_v2"
    },
    {
      "id": "ams-mgp-property-build",
      "phase": "ship",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "New AMS MGP property at /ams/mgp/ + sync-from-legacy bridge + full historical backfill",
      "path": "/z/trading/dev/ams/mgp/",
      "completed_at": "2026-05-21",
      "created_at": "2026-05-21",
      "note": "5/21 16:50-18:16 ET. New brief.php + _content.php + index.php + sync-from-legacy.php under /ams/mgp/ \u2014 third property ported into the AMS shell (after Cockpit + Scorecards). sync-from-legacy.php then copied every legacy MGP page 4/17->5/21 (morning/midday/recap variants, ~150 files + alpha-trend-2026-04-17.html) into stage /trading/ams/mgp/ (17:50-18:02) then dev (18:03-18:16) + manifest + index. AMS MGP archive now has parity with legacy. NOTE: distinct from ams-mgp-fork-prep (the gen_mgp_ams.py generator fork) \u2014 this is the web property + legacy bridge. Still NOT wired to a live entitlement/scope-aware generator. Recovered via supplemental /trading/dev/ams scan.",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "ams-playbooks-system-scaffold",
      "phase": "design",
      "status": "in_progress",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS Playbooks + System property dirs scaffolded (next ports beginning)",
      "path": "/z/trading/dev/ams/playbooks/",
      "created_at": "2026-05-23",
      "note": "New playbooks/ + system/ directories appeared 23:12-23:13 ET 5/23 in both dev /trading/dev/ams/ and stage /trading/ams/ \u2014 likely the start of the next two AMS property ports (Playbooks + System hub) following Cockpit/Scorecards/MGP. Empty scaffold so far, no content files. Tracking as in_progress; flip to completed when source lands.",
      "property": "system_contract",
      "project": "atlas"
    },
    {
      "id": "mgp-v2-audit-response-2026-05-27",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "MGP v2 audit response: tape_read_v2 + sector_cycle_v2 + chiphazzard_trace + phase1_preview",
      "path": "/z/setup_daily_reporting/reports/{tape_read_v2,sector_cycle_v2,chiphazzard_trace,phase1_preview}.py",
      "completed_at": "2026-05-27",
      "created_at": "2026-05-27",
      "property": "mgp",
      "note": "Parallel build of four new modules in response to the 2026-05-27 MGP audit. tape_read_v2 fixes the vocabulary-overload bug (v1 used 'losing 5SMA' to mean both slope-turning-negative and price-below-line) + adds build_net_posture() (5-layer signed vote across futures/daily/intraday/breadth/macro). sector_cycle_v2 always demotes 'clear' confidence to 'tentative' on weak_signal, gates the 'WE ARE HERE' marker + hardcoded blurb behind confidence, renames 'Economic cycle' chip to 'Rotation-implied economic phase'. chiphazzard_trace wraps classify_state with gate_fired/gate_inputs/last_bar_ts/candle_closed/is_partial/is_stale/freshness_note so the MGP can render 'Why this label?' tooltips and validate.py can block publish on stale candles. phase1_preview.py renders v1 vs v2 side-by-side at /mgp/v2/phase1_preview.html. Zero edits to v1 files.",
      "project": "atlas"
    },
    {
      "id": "param-surface-tod-lane-2026-05-29",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Param Surface: NEW Time-of-Day lane (6 intraday buckets + comparison + SD-expansion)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "completed_at": "2026-05-29",
      "created_at": "2026-05-29",
      "property": "param_surface",
      "note": "First time the param surface is sliced by time-of-day. Six buckets: open (9:30-10:00), early (10:00-10:30), mid_am (10:30-11:30), midday (11:30-13:00), mid_pm (13:00-14:30), late (14:30-15:55). Each emits a heatmap + a 3D surface PNG (12 PNGs) plus tod_comparison.png + tod_sd_expansion.png (14 total). Auto-pipeline regen at 16:31-16:32 ET. Tests whether VWAP-band edges have different optimal stop/target SDs at different intraday phases.",
      "project": "strat_lib"
    },
    {
      "id": "mgp-midday-cron-specific-regression-2026-05-28-29",
      "phase": "diagnose",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "title": "MGP MID-DAY edition specifically failed 5/28 + 5/29 (morning + recap deliver, mid-day silently drops)",
      "path": "/z/setup_daily_reporting/reports/",
      "created_at": "2026-05-29",
      "property": "mgp",
      "note": "Failure is narrower than the 5/21-5/22 'whole MGP path' regression: 5/28 broke mid-day AND recap (both missing from notify.log); 5/29 morning (05:41) and recap (17:39) both delivered but mid-day fired ZERO times. That isolates the bug to the mid-day-edition pipeline specifically \u2014 its scheduled task, gen_mgp_midday.py (if separate), or the SMTP call from that runner. ALSO note 5/29 recap was ~37 min later than usual (17:39 vs ~17:00) \u2014 might be same scheduler family. Diagnose before any Vital-drop cutover; the 3-clean-days verification gate is fully broken (5/26+5/27 clean, 5/28 + 5/29 broken).",
      "project": "atlas"
    },
    {
      "id": "nightly-pm-three-day-silent-gap-2026-05-26-to-28",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "title": "Nightly PM: 3 consecutive silent failures (5/26 crashed mid-run, 5/27 + 5/28 no-fire)",
      "path": "/z/logs/nightly-startup.log",
      "created_at": "2026-05-29",
      "property": "pm_system",
      "note": "Longest silent gap yet. 5/26 23:05:47 LAUNCHED is in nightly-startup.log but no DONE \u2014 agent crashed mid-run (per SKILL diagnostic). 5/27 + 5/28 are completely absent \u2014 no LAUNCHED, no DONE \u2014 daemon never fired. 5/29 (tonight) recovered and backfilled all three. Running pattern this month: 5/20/21/22 no-fire, 5/24 no-fire, 5/26 crash-mid-run, 5/27 no-fire, 5/28 no-fire. Eight failures in 10 days. The existing nightly-pm-silent-gap-2026-05-20-to-22 task is now stale \u2014 supersedes it. The wake-from-sleep / auto-login decision can't keep getting deferred.",
      "project": "atlas"
    },
    {
      "id": "taxonomy-reconciliation-audit-2026-06-01",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Strategy taxonomy reconciliation: read-only audit + implementation plan (2 docs)",
      "path": "/z/public_html/system/STRATEGY_TAXONOMY_RECONCILIATION_AUDIT.md",
      "completed_at": "2026-06-01",
      "created_at": "2026-06-01",
      "note": "Two paired /system/ docs shipped. AUDIT: reconciles strategy taxonomy across 5 surfaces (Scorecards _scientific_overlay.js CATALOG, Strategy Library, Strategy Bar playbooks.json, system/strategies.json, DB main.trade_context). Headline: 2026-05-04 handoff ~50% executed (frontend done, DB-side + Strategy Bar never run; main.user_strategies table does not exist). Found 52% of trades (148/283) carry orphan labels (Unclassified 94, Chase/Extended 54) in no canonical source, and system/strategies.json is a divergent fork. PLAN: DDL + seed scripts, orphan-label decision table, root-cause (gen_scorecard.py emits orphans directly + TREND_TRADE relabel bug at line 1467), 10-step execution order. PLAN-ONLY \u2014 awaits James's decisions on plan section 3 before any DB writes / site edits / VCET touches.",
      "project": "atlas"
    },
    {
      "id": "best-may-2026-examples-page",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "Top wins/losses - May 2026 best-trades examples page",
      "path": "/z/public_html/playbooks/best-may-2026/",
      "completed_at": "2026-06-01",
      "created_at": "2026-06-01",
      "note": "Best-trades selector run for May 2026, mirroring the best-april-2026 page. Source data generated under playbook_data/best_may_2026/ (ETHU 5/06, IBIT 5/26, MSFT 5/07, NVDA 5/06+5/27+5/28 minute JSON + per-ticker daily JSON) from may_2026_trades_raw.json + may_2026_sessions_rth.json. Page published at /playbooks/best-may-2026/.",
      "project": "multi_trader"
    },
    {
      "id": "taxonomy-james-decisions-section3",
      "phase": "phase-0",
      "status": "completed",
      "owner": "user",
      "priority": "P1",
      "title": "DECIDE: taxonomy plan section-3 calls (orphan labels + trend-trade-short + relabel bug)",
      "path": "/z/public_html/system/TAXONOMY_RECONCILIATION_PATCH_PLAN.md",
      "completed_at": "2026-06-02",
      "created_at": "2026-06-01",
      "note": "RESOLVED 2026-06-02: James locked all 8 decisions (see PATCH_PLAN 'James's locked decisions' table). Outcomes: (1) Chase/Extended -> NEW sentinel slug chase-extended, family off-book (no alias); (2) Unclassified -> NEW sentinel slug unclassified, family off-book, expectancy_r NULL; (3) Hitchhiker -> NEW canonical slug, family level_breakout (NOT an FL2 alias); (4) Range Break -> new range-break + htf-range-break, opening-range-break retained; (5) add trend-trade-short to CATALOG; (6) fix gen_scorecard.py:1465 TREND_TRADE relabel; (7) strategies.json VERIFIED as live runtime consumer (4 Python readers) -> archival OFF the table, regenerate from main.user_strategies; (8) add canonical_strategy_slug + raw_classifier_output to trade_context. CATALOG grows 19->25 slugs, ALIASES 30->~50. ORIGINAL NOTE: Blocks the entire reconciliation (DB migration, scorecard overlay for the 52% orphan trades, VCET validation). Eight calls awaiting James: (3.1) Chase/Extended -> new sentinel slug [rec] vs alias to fl2-fash-late; (3.2) Unclassified -> sentinel slug + expectancy_r NULL [rec]; (3.3) Hitchhiker -> alias to fl2-fash-late [rec]; (3.4) Range Break -> two-alias disambiguation [rec]; (3.5) add trend-trade-short to CATALOG [rec]; (3.8) fix gen_scorecard.py:1467 TREND_TRADE relabel bug [rec]; (5) confirm whether system/strategies.json has any live consumer; (8.7) add canonical_strategy_slug + raw_classifier_output columns to trade_context [rec].",
      "property": "foundation",
      "project": "atlas"
    },
    {
      "id": "taxonomy-patch-plan-doc-2026-06-02",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Taxonomy reconciliation PATCH PLAN (third /system/ doc) capturing James's locked decisions",
      "path": "/z/public_html/system/TAXONOMY_RECONCILIATION_PATCH_PLAN.md",
      "completed_at": "2026-06-02",
      "created_at": "2026-06-02",
      "property": "foundation",
      "note": "Third doc in the reconciliation set (after AUDIT + IMPLEMENTATION_PLAN). Translates James's 8 locked decisions into a concrete patch plan: gen_scorecard.py constant map (CANONICAL_SLUG/CANONICAL_DISPLAY + slug_for helper), CATALOG growth 19->25 slugs, ALIASES 30->~50 keys, DDL for main.user_strategies, additive trade_context columns (canonical_strategy_slug + raw_classifier_output, dev/staging only). Doc header still reads PLAN-ONLY for DB/SQL, but the gen_scorecard.py code half WAS applied today (see taxonomy-gen-scorecard-patch-2026-06-02). DB writes / SQL execution still pending.",
      "project": "atlas"
    },
    {
      "id": "taxonomy-gen-scorecard-patch-2026-06-02",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Apply gen_scorecard.py taxonomy patch: CANONICAL_SLUG map + TREND_TRADE relabel bug fix",
      "path": "/z/setup_daily_reporting/reports/gen_scorecard.py",
      "completed_at": "2026-06-02",
      "created_at": "2026-06-02",
      "property": "scorecards",
      "note": "Code half of the taxonomy reconciliation landed. Backup taken first (gen_scorecard.py.pre-taxonomy-patch-2026-06-02.bak, 06:29). Patched file (07:17) now carries module-level CANONICAL_SLUG / CANONICAL_DISPLAY maps (line ~129) + slug_for() helper (line ~153) emitting canonical slugs including the new chase-extended and unclassified sentinels (line 486), and the TREND_TRADE relabel bug fix at ~line 1517 (was merging trend-trade into 'vwap-pull-in -> trend'). This stops gen_scorecard.py from minting fresh orphan labels every EOD. Companion SQL seed written (seed_user_strategies_20260602.sql) for main.user_strategies but NOT yet executed against any DB. NOTE: PATCH_PLAN doc still flags PLAN-ONLY; code edit was applied ahead of that doc's status update.",
      "project": "multi_trader"
    },
    {
      "id": "user-strategies-seed-sql-2026-06-02",
      "phase": "phase-0",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Execute seed_user_strategies_20260602.sql to create + populate main.user_strategies (dev/staging)",
      "path": "/z/setup_daily_reporting/reports/seed_user_strategies_20260602.sql",
      "created_at": "2026-06-02",
      "property": "foundation",
      "note": "DDL + seed authored 06:50 today as the DB half of the taxonomy reconciliation (the 2026-05-04 handoff never created this table). SQL is written but NOT yet run. Next: execute against dev/staging only (additive), then add canonical_strategy_slug + raw_classifier_output columns to main.trade_context, then regenerate system/strategies.json from the new table with the direction-suffix translation layer (PATCH_PLAN section 8.4). This is the current unblock now that decisions are locked + gen_scorecard.py is patched.",
      "project": "atlas"
    },
    {
      "id": "nvda-playbook-2026-06-02",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "NVDA playbook published (2026-06-02)",
      "path": "/z/public_html/playbooks/nvda-2026-06-02.html",
      "completed_at": "2026-06-02",
      "created_at": "2026-06-02",
      "property": "playbooks",
      "note": "Template-driven playbook for NVDA published at /playbooks/nvda-2026-06-02.html with companion nvda-2026-06-02-data.js (17:37 ET).",
      "project": "multi_trader"
    },
    {
      "id": "pvp-build-report-refresh-2026-06-02",
      "phase": "phase-scanners",
      "status": "completed",
      "owner": "claude",
      "title": "Pocket-Pivots (PVP) build_report.py + index.html refresh",
      "path": "/z/gt_dev/db_reports/pocket_pivots/build_report.py",
      "completed_at": "2026-06-02",
      "created_at": "2026-06-02",
      "property": "pm_system",
      "note": "PVP Volume-Pockets scanner build_report.py + published index.html updated 14:21 ET. See reference_pvp_pipeline.",
      "project": "atlas"
    },
    {
      "id": "mgp-recap-edition-missed-2026-06-02",
      "phase": "diagnose",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "MGP EOD RECAP edition silently missing \u2014 now 4 consecutive trading days (6/2, 6/3, 6/4, 6/5); last good recap 6/1",
      "path": "/z/setup_daily_reporting/reports/",
      "created_at": "2026-06-02",
      "completed_at": "2026-06-08",
      "property": "mgp",
      "note": "RESOLVED 6/8 \u2014 the recap edition FIRED today at 16:37 ET (notify.log: 'Domino Brief - End of Day Recap 2026-06-08 \u00b7 chop day', sent to all 3 recipients; mgp/2026-06-08-recap.html on disk). First good recap since 6/1, ending the 4-trading-day silent-regression streak (6/2\u20136/5). All three editions delivered clean today: Morning 05:41, Mid-Day 11:39, Recap 16:37. Recap runner is back in service. Original: recap absent from notify.log + disk 6/2\u20136/5; bug was recap-runner-specific.",
      "project": "atlas"
    },
    {
      "id": "trade-context-phase1-6-canonical-slug-2026-06-03",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "scorecards",
      "title": "trade_context Phase 1.6 + canonical_strategy_slug live in main.trade_context",
      "path": "/z/setup_daily_reporting/reports/trade_context.py",
      "created_at": "2026-06-03",
      "completed_at": "2026-06-04",
      "note": "trade_context.py Phase 1.6 (backup .pre-phase-1-6-2026-06-02.bak) + backfill_phase_2_20260603.sql (6/3) + phase_3a_views_20260603.sql (6/4). canonical_strategy_slug is now a live, populated column in main.trade_context \u2014 confirmed because the 6/4 PULL_IN_TO_TREND_MECHANISM_AUDIT queries it directly (slug='vwap-pull-in -> trend', n=15). The DB half of the taxonomy work advanced; main.user_strategies table creation not separately confirmed by the file scan.",
      "project": "atlas"
    },
    {
      "id": "vwap-pull-in-quality-gate-2026-06-04",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "VWAP Pull-In Quality Gate doc set (split diagnostic + gate + cheatsheet + MGP block + checklist)",
      "path": "/z/public_html/system/VWAP_PULL_IN_QUALITY_GATE.md",
      "created_at": "2026-06-04",
      "completed_at": "2026-06-04",
      "note": "GREEN/YELLOW/RED pre-trade gate on RVOL + gap (not entry technique). Sample: 127 plain PI (-$364) + 15 PI->trend (+$367) on main.trade_context thru 6/3. Discriminator is calm vs hot tape: plain-PI losers cluster RVOL>2.5 / gap>1.5% / {IBIT,AMD,ETHU,RGTI}; IBIT alone = -$185 of the -$364. Ships with VWAP_PULL_IN_SPLIT_DIAGNOSTIC, _CHEATSHEET, MGP_BLOCK_VWAP_PULL_IN_QUALITY_GATE, INTRADAY_CHECKLIST_VWAP_PULL_IN, CANONICAL_STRATEGY_PERFORMANCE_REPORT + PULL_IN_RESEARCH_HUB anchor. = B2 (Quality Gate review) from the three-buckets workstream.",
      "project": "classifier"
    },
    {
      "id": "pull-in-to-trend-mechanism-audit-2026-06-04",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "Pull-In -> Trend mechanism audit: bucket is MIXED, edge is rotation scalps not conversions",
      "path": "/z/public_html/system/PULL_IN_TO_TREND_MECHANISM_AUDIT.md",
      "created_at": "2026-06-04",
      "completed_at": "2026-06-04",
      "note": "VCET v1.0 in-memory over all 15 'vwap-pull-in -> trend' trades. Of 6 classifiable: only 1 (JOBY 5/28) is a true conversion; 3 are rotation scalps (+$96, 75.5% of classifiable P&L), 2 wrong-side escapes. 9 insufficient_data (6 NO_BAR older than kti window, +$235). Confirms the outcome-based slug is NOT mechanism-confirmed \u2014 validates the falsified state-at-entry pivot. Ships with VWAP_PULL_IN_CLASSIFIER_REFINEMENT_PLAN + PI_TREND_ROTATION_WINNERS_REVIEW.",
      "project": "classifier"
    },
    {
      "id": "chase-extended-audit-publish-2026-06-04",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "property": "live_classifier",
      "title": "Chase-Extended audit deliverable set published to /system/",
      "path": "/z/public_html/system/CHASE_EXTENDED_AUDIT.html",
      "created_at": "2026-06-04",
      "completed_at": "2026-06-04",
      "note": "Full publish of the chase-extended audit: CHASE_EXTENDED_AUDIT.md/.html + evidence.csv + reclassification.csv + queries.sql. Sentinel-vocab reclassification reference behind the Quality Gate \u00a73 setup-reclassification rules.",
      "project": "classifier"
    },
    {
      "id": "vcet-v1-1-acceptance-2026-06-05",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "property": "live_classifier",
      "title": "VCET v1.1 UNBLOCKED \u2014 all 13 acceptance checks passed (synthetic + real-data)",
      "path": "/z/public_html/system/VCET_V1_1_ACCEPTANCE_SUMMARY.html",
      "created_at": "2026-06-05",
      "completed_at": "2026-06-05",
      "note": "kite_v1.1.0 builds clean, runs end-to-end on a 390-bar multi-regime synthetic dataset, satisfies all 7 v1.0 gates + 6 v1.1 CBO gates (column presence, current-bar exclusion, v1.0 reproducibility under enable_cbo=False, gating, CBO score contribution). Canonical columns present: channel_high/low/mid/width/width_pct + cbo_break_long/short/inside_channel/break_strength. Clears the v1.1-package blocker that froze B1 (VCET engine) since 5/25. Ships with VCET_V1_1_MECHANISM_VALIDATION_PLAN + _ACCEPTANCE_TEST_PLAN + _SOURCE_RECONCILIATION (authored 6/4).",
      "project": "classifier"
    },
    {
      "id": "kti-96-trade-pilot-2026-06-05",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "Stage-A kti 1-min pilot: VWAP/SD recompute validated, VCET v1.1+CBO works on real bars",
      "path": "/z/public_html/system/KTI_96_TRADE_PILOT_RESULTS.html",
      "created_at": "2026-06-05",
      "completed_at": "2026-06-05",
      "note": "62/62 fetched 1-min pairs passed QA, 0 engine errors. 1-min VWAP recompute matches engine 5-min VWAP (median diff 0.023%). VCET v1.1+CBO validated on JOBY 5/28 real data (CBO +10/+25min, width 1.65->5.78, slope +2.3->+51.8deg, state->TREND_UP). STRUCTURAL CAVEAT: kti coverage is age-bounded \u2014 June 100% / May 31.6% / April 0% (143 trades NO_BARS). Ships with VWAP_SD_RECOMPUTE_METHOD + LIBERATOR_CLOUDQUANT_SOURCE_VALIDATION (the April-gap fill path) + ENTRY_FEATURE_SAMPLE_OUTPUT.csv + UNCOMPUTABLE_TRADES_LIST.csv.",
      "project": "classifier"
    },
    {
      "id": "auto-pipeline-2026-06-06",
      "phase": "phase-ops",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "property": "param_surface_db",
      "title": "Quiet Saturday \u2014 auto-pipelines clean (scorecard + full param-surface PNG regen incl TOD surfaces)",
      "path": "/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface-db/",
      "created_at": "2026-06-06",
      "completed_at": "2026-06-06",
      "note": "Market closed (Sat). No MGP editions expected/fired \u2014 correctly absent. Auto-pipeline output only, all clean at 16:30 ET: scorecards/trade_scorecard_2026-06-06.html + manifest; full james-monthly-reports rebuild (June index + vwap-bands + param-surface + param-surface-db) with complete gen_surface_pngs.py PNG set INCLUDING the time-of-day surfaces (tod_surface/heatmap open/early/mid_am/midday/late + tod_comparison + tod_sd_expansion + CROX); leave_on_table_archive/2026-06-05.json. No source-code changes detected (HUD/dev subtree scan stalled on WebDAV again; weekend). NOTE: [WATCHDOG] RED: todos_snapshot fired 06:31 ET \u2014 recurring freshness regression (also 6/2, 6/3, 6/5).",
      "project": "atlas"
    },
    {
      "id": "liberator-datasource-blocker-2026-06-07",
      "phase": "diagnose",
      "status": "blocked",
      "owner": "claude",
      "priority": "P0",
      "property": "live_classifier",
      "title": "Liberator minute_bars entitlement LOST (27->10 datasets) \u2014 Stage B mechanism backfill now BLOCKED on data",
      "path": "/z/public_html/system/DATA_SOURCE_RECOVERY_AND_STAGE_B_FALLBACK_PLAN.html",
      "created_at": "2026-06-07",
      "note": "Sunday research day. The 20-pair Liberator/kti parity test (Stage A.1) could NOT run: kti liberator query --dataset minute_bars times out 'context deadline exceeded' on every pair because the account's Liberator entitlements dropped from 27 datasets to 10 between 6/5 15:00 and 6/7 19:55 ET \u2014 only fis_*/s3_partners/wonda remain, NO market_data (no minute_bars). The one cached pair (NVDA 6/2) showed EXACT parity ($0.0000 close diff, 390/390 bars) so the source is trustworthy; the blocker is purely access. Recovery plan: send CloudQuant a minute_bars restoration request FIRST; do NOT lock in a kti-only 30%-coverage Stage B (recent-biased, won't survive review); F5 hybrid (kti recent + Liberator older) recommended once access returns. Quality Gate stays primary trading rule; mechanism work pauses if restoration delayed. Ships with LIBERATOR_KTI_PARITY_20_PAIR_REPORT + LIBERATOR_CLOUDQUANT_SOURCE_VALIDATION + LIBERATOR_TIMESTAMP_NORMALIZATION + VWAP_SD_RECOMPUTE_METHOD + PREMARKET_POLICY_RECOMMENDATION (EXCLUDE premarket -> insufficient_data_premkt).",
      "project": "classifier"
    },
    {
      "id": "you-vs-you-scorecard-2026-06-07",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "scorecards",
      "title": "New 'You vs You' weekly self-referential scorecard (generator + page + EOD batch + nav)",
      "path": "/z/setup_daily_reporting/reports/you_vs_you.py",
      "created_at": "2026-06-07",
      "completed_at": "2026-06-07",
      "note": "you_vs_you.py aggregates last N days of an account's fills into one synthesis page: trajectory (improving vs spiraling), the hold-time leak, Good-James vs Bad-James cross-cuts (symbols/time-of-day/sides), behavioral flags, auto-generated marching orders \u2014 benchmarked against the trader's OWN history only, no peer comparison (commission $0.01/share, matches gen_scorecard.py). Published scorecards/you-vs-you.html; wired into scorecards/index.html + rules-card.html; run_all_eod.bat updated to include it in the EOD batch.",
      "project": "multi_trader"
    },
    {
      "id": "auto-pipeline-2026-06-07",
      "phase": "phase-ops",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "property": "param_surface_db",
      "title": "Quiet Sunday market \u2014 auto-pipelines clean (scorecard + full param-surface PNG regen incl TOD surfaces)",
      "path": "/gt_dev/db_reports/james-monthly-reports/",
      "created_at": "2026-06-07",
      "completed_at": "2026-06-07",
      "note": "Market closed (Sun) \u2014 no MGP editions expected/fired (correctly absent; recap miss-count holds at 4). Auto-pipeline clean at 16:30 ET: scorecards/trade_scorecard_2026-06-07.html + manifest; full james-monthly-reports rebuild (June index + vwap-bands + param-surface + param-surface-db) with complete PNG set INCLUDING the time-of-day surfaces (tod_surface/heatmap open/early/mid_am/midday/late + tod_comparison + tod_sd_expansion + CROX). HUD/dev subtree scan ran slow but completed \u2014 only auto-generated noise (error_log/php_error.log/vendor/.bak), no HUD source changes.",
      "project": "atlas"
    },
    {
      "id": "trader-strategy-attribution-report-2026-06-08",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "monthly_reports",
      "title": "Trader Strategy Attribution report published \u2014 60-day, all-6-trader strategy match",
      "path": "/gt_dev/db_reports/james-monthly-reports/strategy-attribution/index.html",
      "created_at": "2026-06-08",
      "completed_at": "2026-06-08",
      "note": "Published strategy-attribution/index.html + analysis.json (generated 14:51 ET; window 2026-04-09\u21922026-06-08, 19,704 trades across 6 traders, total net +$2,999, 86.2% has-VWAP). Per-trader strategy/regime breakdown via the Liberator-bars\u2192VWAP-recon\u2192classify_v2 pipeline. Strategy mix confirms the prior finding: off-book dominates (13,071 of 19,704) and regime fires mostly rotation (16,361) vs trend (617). See project_trader_strategy_report \u2014 sensitive, behind smb.a1prospect.com access, not on the public site.",
      "project": "atlas"
    },
    {
      "id": "auto-pipeline-2026-06-08",
      "phase": "phase-ops",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "property": "param_surface_db",
      "title": "Trading day 6/8 \u2014 all 3 MGP editions clean, recap restored, auto-pipelines clean",
      "path": "/gt_dev/db_reports/james-monthly-reports/",
      "created_at": "2026-06-08",
      "completed_at": "2026-06-08",
      "note": "Monday trading day. MGP delivered all three editions (Morning 05:41 'risk-on \u00b7 breadth -1', Mid-Day 11:39 'yellow lights', Recap 16:37 'chop day') \u2014 the recap firing ends the 4-day regression. Auto-pipeline clean 16:30 ET: scorecards/trade_scorecard_2026-06-08.html + manifest + you-vs-you.html (the new self-referential scorecard renders in the EOD batch); full james-monthly-reports rebuild (June index + vwap-bands + param-surface + param-surface-db) with complete TOD PNG set + CROX; vwap-research/index.html refreshed 16:46; leave_on_table_archive/2026-06-05.json. NOTE: a config edit landed at 09:04 ET in gt_dev/db_reports/smb-reports-v2/accountability/config.json (backup config-before-edit-2026-06-08T13-04-54Z.json written) \u2014 not a nightly-PM change; flagging for visibility. HUD/dev source subtree scan returned EMPTY (no accountability-v2 source changes).",
      "project": "atlas"
    },
    {
      "id": "vwap-pull-in-gate-validation-2026-06-09",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "VWAP Pull-In Quality Gate \u2014 walk-forward OOS validation + gate-validation audit + RVOL/instrument-type crosstab",
      "path": "/z/public_html/system/VWAP_PULL_IN_WALKFORWARD_OOS_VALIDATION.html",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Read-only validation add-on (no DB writes) over the vwap-band-entry-surface Liberator backfill. Time-split (train 4/10-5/13 n=77 / test 5/14-6/4 n=55, no shuffle) of plain vwap-pull-in. HEADLINE: MIXED \u2014 DO NOT PROMOTE ON THIS OOS ALONE, forward-test required. What HELD OOS (sign-consistent both halves): cut the post-entry fade-exit (most robust; fade 8.7% win in test), pass |gap|>1.5% RED, avoid non-megacap types, and the FULL combined stack (megacap+controlled-gap+cut-fade) stayed positive both halves (but test n=7, underpowered). What FLIPPED: 'megacap=profitable' (+7.05 train -> -2.83 test = relative/least-bad filter only) and entry-gate-ALONE (+12.23 -> -2.95). Test half was a much worse regime (baseline +91 train vs -389 test) and one IBIT 6/3 trade (-$216) is >55% of the test loss. Ships with VWAP_PULL_IN_GATE_VALIDATION_AUDIT.html + VWAP_PULL_IN_RVOL_INSTRUMENT_TYPE_CROSSTAB.html (+ their JSON summaries). Production untouched. Resolves the project_profitability_diagnosis 'no validated OOS edge / walk-forward overfit' question: durable core is the post-entry overlay, not the entry gate alone.",
      "project": "classifier"
    },
    {
      "id": "vwap-pull-in-forward-test-tracker-2026-06-09",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "VWAP Pull-In forward-test tracker + schema + actionable playbook + strategy-library draft",
      "path": "/z/public_html/system/VWAP_PULL_IN_FORWARD_TEST_TRACKER.html",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Operationalizes the walk-forward verdict ('HOLD historical promotion, forward-test next'). Built the forward-test tracker page + vwap_pull_in_forward_test_schema.json (the live log structure), VWAP_PULL_IN_ACTIONABLE_PLAYBOOK.html + vwap_pull_in_actionable_gate.json (the GREEN-setup rules to trade behind the gate: cut the fade, pass gap-RED, avoid bad type), and vwap_pull_in_strategy_library_draft.json (the candidate Strategy Library entry, NOT yet promoted). Next working day = start logging live GREEN setups against this tracker.",
      "project": "classifier"
    },
    {
      "id": "backside-classifier-aaoi-audit-2026-06-09",
      "phase": "diagnose",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "property": "live_classifier",
      "title": "Backside-Long classifier \u2014 AAOI-triggered lifecycle rebuild + 90-day read-only backtest",
      "path": "/z/trading/dev/backside-classifier-audit/BACKSIDE-CLASSIFIER-AUDIT.md",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Read-only research (NOT web-published, NOT promoted). AAOI trade triggered a rebuild of Backside-Long as a setup/outcome-separated lifecycle: backside_lifecycle.py, AAOI single-name audit (audit_aaoi.py + validate_aaoi.py), then a full bar pull (kti minute bars 3/10-6/8, extended back to 3/10) and TWO backtests \u2014 backtest.py (recent window) + backtest90.py (90-day window) \u2014 each with an HTML report (backtest-report.html, backtest90-report.html). The old gate tested entry DEPTH not the VWAP reclaim, so reclaim entries were mislabeled. STILL awaiting James's reclaim-ceiling calibration before any promotion (P1-P6). See project_backside_classifier_rebuild.",
      "project": "classifier"
    },
    {
      "id": "scorecard-chart-viewer-2026-06-09",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "property": "scorecards",
      "title": "Scorecard chart-viewer module (chart_viewer.js + v2 + tooltip) + briefing",
      "path": "/z/public_html/_shared/chart_viewer_v2.js",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "New shared chart-viewer for the scorecard surface: _shared/chart_viewer.js then _shared/chart_viewer_v2.js (iteration), plus _shared/tooltip.js. CHART_VIEWER_BRIEFING.md authored in setup_daily_reporting/reports as the spec. Renders per-trade charts to the James chart-defaults (black hollow candles, MA stack, RSI sub-chart) inside scorecard cards.",
      "project": "multi_trader"
    },
    {
      "id": "auto-pipeline-2026-06-09",
      "phase": "phase-ops",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "property": "param_surface_db",
      "title": "Auto-pipelines clean \u2014 all 3 MGP editions + scorecard + you-vs-you + full param-surface regen",
      "path": "/gt_dev/db_reports/james-monthly-reports/",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Trading day (Tue). ALL THREE MGP editions fired clean: Morning 05:41 ('risk-on - breadth +1 - constructive tape'), Mid-Day 11:38 ('thesis broken'), Recap 16:37 ('reversal day') \u2014 2nd consecutive day with a good recap since the runner fix, regression streak stays closed. Auto-pipeline 16:30-16:32: scorecards/trade_scorecard_2026-06-09 + manifest + you-vs-you.html; full james-monthly-reports rebuild (June index + vwap-bands + param-surface + param-surface-db) with complete TOD PNG set + CROX; leave_on_table_archive/2026-06-09.json; vwap-research/index.html refreshed 16:46.",
      "project": "atlas"
    },
    {
      "id": "projected-5dma-environment-buildout-2026-06-09",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "live_classifier",
      "title": "Projected 5DMA Environment \u2014 full research pipeline -> archetype-aware report + live forward-test rig",
      "path": "/z/trading/dev/projected-5dma-environment/PROJECTED_5DMA_ENVIRONMENT_RESEARCH_REPORT.html",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Read-only research (dev only), big build-out of the 6/9 projected-5DMA memo into a working rig. Pulled ~200 symbols of daily bars (pull_daily_bars.py), built projected_ma.py + build_features.py + analyze.py + multitrader analysis, then an ARCHETYPE INTERPRETER (strategy_archetype_map.v1.json + archetype_interpreter.py + archetype_env_rule_matrix.csv) that operationalizes the key finding \u2014 the env signal is archetype-CONDITIONAL (continuation wants up-env, Backside-Long wants the opposite washout). Deliverables: PROJECTED_5DMA_ENVIRONMENT_AUDIT.md, RESEARCH_REPORT.html, STRATEGY_AWARE_ADDENDUM.html, IMPLEMENTATION_PLAN.md, FORWARD_TEST_PLAN_5DMA.md + LIVE_FORWARD_TEST_AND_SCANNER_PLAN.md, plus a live forward-test rig (live_forward_test.py + forward_test_logger.py + PROJECTED_5DMA_FORWARD_TEST_DASHBOARD.html) and an index.html. Recommendation stands: context tag + risk tier, not a trigger. See project_projected_5dma_environment.",
      "project": "classifier"
    },
    {
      "id": "hotkey-engine-buildout-2026-06-09",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "sizing",
      "title": "Setup-Grade Risk Hotkey \u2014 target engine + Kore Backend A + GamePlan export contract",
      "path": "/z/trading/dev/hotkey-engine/SPEC.md",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Read-only/forward-test-only build-out of the Gr8Trade setup-grade hotkey (project_setup_hotkey). Advanced it from 'awaiting' to a fuller forward-test rig: GamePlan EMIT contract now exists (gameplan_export.py + setups.today.sample.json + contract/setups.schema.json), target_engine.py + test (grade->target sizing / auto-stop), Backend A Kore wired (kore_backend.py + config sample + test_kore_backend.py + kore_strategy_skeleton.py), engine.py + hotkey.py front-end, register_speed_key.ps1 for the Gr8Trade speed key, SPEC.md + FORWARD-TEST-CHECKLIST.md. Execution still inert/forward-test pending firm-confirm of Gr8Script/Kore calls. tickets-2026-06-09.jsonl + test-kore.jsonl logged.",
      "project": "atlas"
    },
    {
      "id": "scorecards-multitrader-migration-2026-06-09",
      "phase": "build",
      "status": "in_progress",
      "owner": "claude",
      "priority": "P2",
      "property": "scorecards",
      "title": "Scorecards multi-trader migration \u2014 extract shared core generator",
      "path": "/z/trading/dev/scorecards-multitrader/MIGRATION-NOTES.md",
      "created_at": "2026-06-09",
      "note": "NEW dev dir. Refactor of the scorecard generator toward multi-trader: gen_scorecard_core.py (extracted shared core) + gen_scorecard_mt.py (multi-trader driver) + _verify.py + MIGRATION-NOTES.md + _genrun.log. In progress, dev only \u2014 not yet wired into the live scorecards pipeline.",
      "project": "multi_trader"
    },
    {
      "id": "pbs-pvp-pipeline-edits-2026-06-09",
      "phase": "phase-ops",
      "status": "completed",
      "owner": "claude",
      "priority": "P3",
      "property": "param_surface_db",
      "title": "PBS/PVP pipeline \u2014 HA script edits + clean live/build runs",
      "path": "/z/trading/dev/pbs/ha/SCHEDULE.md",
      "created_at": "2026-06-09",
      "completed_at": "2026-06-09",
      "note": "Morning edits to the PBS HA layer (pbs/ha/run_if_stale.sh, PbsTime.pm, _pbs_time.sh, pbs_live.sh, run_pbs_*.cmd, CLAUDE.md, SCHEDULE.md) and PVP live (pvp_live_tick.pl, pvp_live.sh, pvp-charts.js). PVP live + build ran clean (pp_swing_real/pp_investor_real, state history pp_*_2026-06-09, universe_pp_daily gz, last_build.json 20:34). Scheduled-pipeline housekeeping, not a feature.",
      "project": "atlas"
    },
    {
      "id": "hotkey-engine-execution-layer-2026-06-10",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Setup-Grade Hotkey: execution/engine layer built (engine + move-stop + daily-stop wiring + deploy)",
      "path": "/z/trading/dev/hotkey-engine/engine.py",
      "completed_at": "2026-06-10",
      "created_at": "2026-06-10",
      "note": "Big build day on the Setup-Grade Risk Hotkey (forward-test only, real execution still gated). Shipped engine.py + hotkey.py, move_stop.py + test_move_stop.py, daily_stop_source.py + set_daily_stop.py (daily-stop integration), risk_profile.py + risk_profile.json, make_today_setups.py + setups.today.json, test_overrides.py, register_speed_key.ps1, deploy_local.ps1 (local deploy), forward_test_report.py, HANDOFF-einstein-box.md, and refreshed SPEC.md + SPEC-VWAP-MODES.md. Advances project_setup_hotkey from 'awaiting firm-confirm' toward a runnable local forward-test rig. Next move = RUN the forward test, not build more.",
      "property": "alerts-hotkeys-etc",
      "project": "atlas"
    },
    {
      "id": "projected-5dma-phase1-operationalize-2026-06-10",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "Projected 5DMA Environment: Phase 1 operator pack + today's review + forward-test dashboard",
      "path": "/z/trading/dev/projected-5dma-environment/index.html",
      "completed_at": "2026-06-10",
      "created_at": "2026-06-10",
      "note": "Operationalized the 6/9 rig (dev/read-only). Shipped PHASE1_DAILY_OPERATOR_GUIDE.md + PHASE1_COMMAND_EXAMPLES.md + PHASE1_REVIEW_CHECKLIST.md, the TODAY_PHASE1_5DMA_REVIEW_2026-06-10.html daily review, PROJECTED_5DMA_FORWARD_TEST_DASHBOARD.html, scripts/ + _data/ + _out/, and index.html. Recall the archetype-conditional finding: continuation wants up-env, Backside-Long wants the OPPOSITE (bearish washout). Rec stays context-tag + risk-tier, not a trigger. Another forward-test rig that now needs to be RUN.",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "backside-promotion-package-2026-06-10",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "Backside-Long classifier: PROMOTION-PACKAGE.md authored + safety rig (still awaiting James calibration)",
      "path": "/z/trading/dev/backside-classifier-audit/PROMOTION-PACKAGE.md",
      "completed_at": "2026-06-10",
      "created_at": "2026-06-10",
      "note": "Continued the AAOI-triggered rebuild (dev/read-only). Wrote PROMOTION-PACKAGE.md, gen_safety.py + safety_data.json, refreshed backside_lifecycle.py, and a proposed/ folder. This packages the rebuilt setup/outcome-separated lifecycle for promotion review. NOT promoted: still blocked on the single James decision (the reclaim ceiling) before the P1-P6 path starts. The package authoring means it is now truly one calibration call from promotion.",
      "property": "scorecards",
      "project": "atlas"
    },
    {
      "id": "vcet-replay-viewer-smci-mrvl-2026-06-11",
      "phase": "build",
      "status": "completed",
      "priority": "P2",
      "title": "VCET Replay Viewer (Phase 1) \u2014 SMCI + MRVL 6/11 visual replays + comparison + consistency audit + viewer index + v1.1 overlay usage guide",
      "owner": "James",
      "property": "system",
      "note": "Reusable symbol+date replay tool, Phase 1 of replay->library->scanner. Today: SMCI 6/11 visual replay + postmortem (110-sh fat-finger Max-Size Guard class; +$501 model vs -$54.59 actual), MRVL 6/11 visual replay, SMCI<->MRVL comparison, replay viewer index, cross-replay consistency audit, and VCET v1.1 chart-overlay usage guide. Real karb 1-min bars, inferred VCET labels.",
      "path": "/system/VCET_REPLAY_VIEWER_INDEX.html",
      "completed_at": "2026-06-11",
      "project": "atlas"
    },
    {
      "id": "auto-pipeline-2026-06-11",
      "phase": "ship",
      "status": "completed",
      "priority": "P3",
      "title": "Trading day 6/11 \u2014 all 3 MGP editions + MGP v2 clean, scorecard + you-vs-you + full param-surface PNG regen",
      "owner": "system",
      "property": "mgp",
      "note": "Morning 05:41, Mid-Day 11:38/11:42, Recap 16:37/16:44 all emailed (chop day). Scorecard 2026-06-11 + you-vs-you regenerated; full vwap-bands param-surface incl TOD surfaces regenerated; vwap-research index refreshed. HUD accountability config.json edited via cockpit UI (config-before-edit backups 09:17) \u2014 cockpit self-edit, not source change.",
      "path": "/mgp/2026-06-11-recap.html",
      "completed_at": "2026-06-11",
      "project": "atlas"
    },
    {
      "id": "vcet-trade-readiness-entry-matrix-2026-06-12",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET Trade Readiness & Entry Matrix \u2014 SMCI + MRVL across 8 candidate entry rules",
      "path": "/z/public_html/system/VCET_TRADE_READINESS_AND_ENTRY_RULE_STUDY_SMCI_MRVL.html",
      "completed_at": "2026-06-12",
      "created_at": "2026-06-12",
      "note": "Per-symbol Trade-Readiness & Entry matrices for SMCI and MRVL (6/11) plus the combined 8-candidate-entry-rule study answering 'when did this turn from good-idea/bad-trade into an actionable trade'. Published HTML + JSON for each. Step 1 of the VCET canonical next-steps list.",
      "project": "atlas"
    },
    {
      "id": "vcet-replay-canonical-reconciliation-2026-06-12",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET Replay \u2014 consistency audit + canonical status/next-steps + SMCI full-order-log postmortem",
      "path": "/z/public_html/system/VCET_REPLAY_CANONICAL_STATUS_AND_NEXT_STEPS.html",
      "completed_at": "2026-06-12",
      "created_at": "2026-06-12",
      "note": "Reconciled all replay-viewer work: one canonical generator + template, all 7 live URLs HTTP 200 with on-disk bytes = served bytes, duplicates in gt_dev\\jd labeled (nothing deleted). SMCI 'actual' P&L upgraded from the FULL 54-fill order log: day -$117.83; 12:24 event entry was 100 sh wavg 29.570 (not 200 @ 29.5725); event campaign -$16.79 replaces the +$7.20 residual-flatten estimate. Consistency audit + postmortem published. Honesty contract intact: VCET state zones still INFERRED (v1.1 engine never ran); gate/RVOL declared not-computed.",
      "project": "atlas"
    },
    {
      "id": "vcet-spcx-expansion-failure-replay-2026-06-12",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VCET Replay \u2014 SPCX 6/12 same-day expansion-failure replay (3rd entry; first same-day) + viewer index refresh",
      "path": "/z/public_html/system/SPCX_20260612_VCET_EXPANSION_FAILURE_REPLAY.html",
      "completed_at": "2026-06-12",
      "created_at": "2026-06-12",
      "note": "Added SPCX 2026-06-12 as the third replay-library entry and the first same-day case, capturing a VCET expansion-failure pattern (real karb 1-min bars + computed bands/CBO, zones inferred). Viewer index updated to list it. Advances the Phase 2 batch-replay-library path toward scanner-readiness.",
      "project": "atlas"
    },
    {
      "id": "vcet-indicator-strategy-spec-v0-2026-06-12",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VCET Indicator Strategy Spec v0 + today's trade guardrail review (6/12)",
      "path": "/z/public_html/system/VCET_INDICATOR_STRATEGY_SPEC_V0.html",
      "completed_at": "2026-06-12",
      "created_at": "2026-06-12",
      "note": "First-cut VCET indicator/strategy spec (v0) plus a same-day guardrail review of 6/12 trades against the VCET rules. Research/spec layer; no classifier or taxonomy change, no Stage B engine run.",
      "project": "atlas"
    },
    {
      "id": "stocks-in-play-top5-generator-refresh-2026-06-12",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Stocks-In-Play / Top-5 setups + MGP v2 universe generator refresh",
      "path": "/z/setup_daily_reporting/reports/top5_trade_setups.py",
      "completed_at": "2026-06-12",
      "created_at": "2026-06-12",
      "note": "Evening edits to the pre-market universe pipeline: stocks_in_play_universe.py, stocks_in_play_v2.py, top5_trade_setups.py, gen_mgp_v2.py, and _shared/tooltip.js. Universe caches (pvp/pbs) regenerated. Supports the MGP v2 parallel build.",
      "project": "atlas"
    },
    {
      "id": "vcet-phase-0a-closeout-2026-06-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET Phase 0A baseline CLOSED \u2014 v0.1.1 ruleset locked",
      "path": "/system/VCET_PHASE_0A_CLOSEOUT.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Single-knob expansion gate cut bleed 57% (-$87k->-$38k) and held OOS; SMCI alone profitable (+$163/50% win). v0.1.1 ruleset locked. Closeout page published; Phase 0b (scanner filter) opened.",
      "project": "atlas"
    },
    {
      "id": "vcet-phase-0a-backtest-harness-2026-06-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VCET Phase 0A backtest harness \u2014 architecture, runner spec, scorecard wireframe + smoke/dryrun audits",
      "path": "/system/VCET_PHASE_0A_BACKTEST_ARCHITECTURE.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Backtest architecture + runner spec + scorecard wireframe authored; smoke-test audit then bigger-dryrun audit run and published (JSON + HTML) to validate the harness before closeout.",
      "project": "atlas"
    },
    {
      "id": "vcet-smci-signal-miss-diagnostic-2026-06-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VCET SMCI 6/11 12:24 signal-miss diagnostic",
      "path": "/system/VCET_SMCI_20260611_1224_SIGNAL_MISS_DIAGNOSTIC.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Root-caused why the SMCI 12:24 conversion was not flagged by the v0.1.1 gate; JSON + HTML diagnostic published to feed the Phase 0B entry contract.",
      "project": "atlas"
    },
    {
      "id": "vcet-phase-0b-entry-contract-2026-06-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET Phase 0B entry contract + proxy-scanner lift",
      "path": "/system/VCET_PHASE_0B_ENTRY_CONTRACT.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Phase 0B opened: entry-gate contract defined and proxy-scanner lift measured (JSON + HTML). Sets the scope for the Phase 0B scanner-filter build (next).",
      "project": "atlas"
    },
    {
      "id": "vcet-phase-0e-param-surface-2026-06-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VCET Phase 0E param surface (stop/target sweep)",
      "path": "/system/VCET_PHASE_0E_PARAM_SURFACE.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Phase 0E param-surface sweep published (JSON + HTML) for VCET stop/target calibration.",
      "project": "atlas"
    },
    {
      "id": "param-surface-monthly-regen-2026-06-14",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "VWAP-bands param-surface monthly report regenerated",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/index.html",
      "completed_at": "2026-06-14",
      "created_at": "2026-06-14",
      "property": "param-surface-db",
      "note": "Per-symbol + combined param-surface/heatmap PNGs and time-of-day surfaces regenerated (CROX/QBTS/BTDR/ALOY + 6 ToD buckets); monthly index refreshed.",
      "project": "atlas"
    },
    {
      "id": "vcet-phase-0e-ema-variant-sweep-2026-06-15",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET Phase 0E param surface expanded \u2014 EMA4/9/20 trail variants + PARTIAL80 (80/20) exit sweeps",
      "path": "/system/VCET_PHASE_0E_PARAM_SURFACE_EMA20.html",
      "completed_at": "2026-06-15",
      "created_at": "2026-06-15",
      "property": "param-surface-db",
      "note": "SHIPPED 2026-06-15 (Mon). Extended yesterday's single 0E stop/target sweep into a multi-variant surface: EMA4 + EMA20 trail sweeps (07:42) then EMA9_PARTIAL80 + EMA20_PARTIAL80 (09:24) adding the 80/20 partial-exit policy as a second axis. Four JSON + four HTML published to /system/VCET_PHASE_0E_PARAM_SURFACE_*. This directly serves the open today_focus item \u2014 the surface now spans MA period (4/9/20) \u00d7 exit policy (full vs 80/20), which is what a stop/target/trail cell gets picked FROM. No cell chosen yet (next step).",
      "project": "atlas"
    },
    {
      "id": "multi-tf-stack-refresh-2026-06-15",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Multi-TF Index Stack v4 + supporting report scripts ran/refreshed for 6/15",
      "path": "/z/setup_daily_reporting/reports/gen_multi_tf_stack.py",
      "completed_at": "2026-06-15",
      "created_at": "2026-06-15",
      "property": "mgp",
      "note": "Multi-TF stack v4 pipeline ran clean for 2026-06-15: chiphazzard_v4.py (stage classifier, 13:35), gen_multi_tf_stack.py (14:08), market_sentiment.py (15:30), dale_vwap_state.py (10:56) iterated; per-symbol 5m+1D caches rebuilt for SPY/QQQ/IWM/SMH/VIX/UVXY; /system/multi-tf-stack/ page refreshed (13:50). _shared/tooltip.js shared UI helper updated (14:28).",
      "project": "atlas"
    },
    {
      "id": "param-surface-monthly-regen-2026-06-15",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "VWAP-bands param-surface monthly report regenerated (daily)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/index.html",
      "completed_at": "2026-06-15",
      "created_at": "2026-06-15",
      "property": "param-surface-db",
      "note": "Daily regen 2026-06-15 (16:37\u201316:38): per-symbol + combined param-surface/heatmap PNGs (CROX/QBTS/BTDR/AMZN/ALOY) and full ToD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion. james-monthly-reports june-2026 + vwap-bands + param-surface-db indexes rebuilt; vwap-research index refreshed (16:46).",
      "project": "atlas"
    },
    {
      "id": "scorecards-daily-2026-06-15",
      "phase": "phase-scorecards",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Daily scorecard + you-vs-you regen for 6/15",
      "path": "/scorecards/trade_scorecard_2026-06-15.html",
      "completed_at": "2026-06-15",
      "created_at": "2026-06-15",
      "property": "scorecards",
      "note": "Daily scorecard pipeline ran clean: trade_scorecard_2026-06-15.html (16:31) + manifest.json + you-vs-you.html (16:39) regenerated.",
      "project": "multi_trader"
    },
    {
      "id": "strategy-views-3view-consolidation-2026-06-16",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Strategy Views Consolidation: 3 views generated from canonical strategies.json master",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/scripts/build_strategies_page.py",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "strategy_library",
      "note": "SHIPPED 2026-06-16 (Tue). Closes the open 'generate 3 views' item from project_strategy_views_consolidation. Built the full canonical pipeline: extract_strategies_master.js (12:42) extracts the master from the library CATALOG; merge_canonical.py (14:48) merged into strategies/strategies.json (3 timestamped .bak snapshots 12:43/14:05/14:36 preserved); then the 3 view generators \u2014 gen_strategy_bar_playbooks.py (15:17 \u2192 strategy-bar/playbooks.json), gen_strategy_registry.py (15:22 \u2192 strategy-registry/index.html), build_strategies_page.py (15:41 \u2192 strategies/index_master_preview.html). Reconciliation: build_strategy_reconciliation.py (13:47 \u2192 /system/strategy_reconciliation_20260616_134701). Supporting SQL/seed scaffold: gen_playbooks_json.py, seed_playbooks_from_json.py, apply_sql.py, apply_safe_first.py, apply_review_step.py. _shared/tooltip.js + tooltip-test.html refreshed (16:27\u201316:30). App-status watchdog went RED on strategies/strategy_bar/strategy_registry at 12:01 mid-rebuild and RECOVERED at 15:31 once the views landed \u2014 expected during the migration.",
      "project": "strat_lib"
    },
    {
      "id": "vcet-cbo-context-canonical-validation-2026-06-16",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET CBO-context canonical validation + Phase 0A quality rerun + 2m CBO context page audit",
      "path": "/system/VCET_CBO_CONTEXT_CANONICAL_VALIDATION.html",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "param-surface-db",
      "note": "SHIPPED 2026-06-16 (Tue). VCET CBO-context validation pass: VCET_CBO_CONTEXT_CANONICAL_VALIDATION (JSON+HTML, 13:44), VCET_PHASE_0A_CBO_CONTEXT_QUALITY_RERUN (JSON+HTML, 13:51), and per-day 2-min CBO-context page audits VCET_20260615/20260616_2M_CBO_CONTEXT_PAGE_AUDIT (09:03 / 12:21). VCET charts regenerated for 6/15 and 6/16 across the full TF set (1m/2m/5m/10m/30m/65m, 12:38). Validates the CBO break-direction\u00d7timing context layer feeding the 0B entry contract.",
      "project": "atlas"
    },
    {
      "id": "param-surface-monthly-regen-2026-06-16",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "VWAP-bands param-surface monthly report regenerated (daily)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/index.html",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "param-surface-db",
      "note": "Daily regen 2026-06-16 (16:38\u201316:39): per-symbol + combined param-surface/heatmap PNGs (CROX/QBTS/BTDR/AMZN/ALOY/MSFT) and full ToD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion. james-monthly-reports june-2026 + vwap-bands + param-surface-db indexes rebuilt; vwap-research index refreshed (16:46).",
      "project": "atlas"
    },
    {
      "id": "scorecards-daily-2026-06-16",
      "phase": "phase-scorecards",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Daily scorecard + you-vs-you regen for 6/16",
      "path": "/scorecards/trade_scorecard_2026-06-16.html",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "scorecards",
      "note": "Daily scorecard pipeline ran clean: trade_scorecard_2026-06-16.html (16:31) + manifest.json + you-vs-you.html (16:40) regenerated.",
      "project": "multi_trader"
    },
    {
      "id": "multi-tf-stack-refresh-2026-06-16",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Multi-TF Index Stack v4 + supporting report scripts ran/refreshed for 6/16",
      "path": "/z/setup_daily_reporting/reports/gen_multi_tf_stack.py",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "mgp",
      "note": "Multi-TF stack v4 ran clean for 2026-06-16: per-symbol 1D caches built pre-market (04:08) + 5m source caches refreshed post-close (19:59) for SPY/QQQ/IWM/SMH/VIX/UVXY; gen_multi_tf_stack.py iterated (16:07); market_sentiment.py (16:34) + temporal_curve_cache; /system/multi-tf-stack/ page refreshed.",
      "project": "atlas"
    },
    {
      "id": "pvp-pocket-pivots-build-2026-06-16",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Pocket-Pivot (PVP) scanner report rebuilt for 6/16",
      "path": "/z/gt_dev/db_reports/pocket_pivots/build_report.py",
      "completed_at": "2026-06-16",
      "created_at": "2026-06-16",
      "property": "param_surface",
      "note": "PVP pocket-pivots scanner rebuilt 2026-06-16 (17:56): build_report.py iterated + /pocket_pivots/index.html regenerated. App-status watchdog flagged GAP_MA RED at 17:31 and RECOVERED at 20:01.",
      "project": "strat_lib"
    },
    {
      "id": "strategies-promote-live-2026-06-17",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Strategy views: canonical master promoted to LIVE /strategies/index.html",
      "path": "/z/public_html/strategies/index.html",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "strategy_library",
      "note": "Follows the 6/16 3-view consolidation. build_strategies_page.py iterated (09:15 -> strategies/index_master_preview.html), strategies/strategies.json re-merged (16:21), then the master preview promoted to live strategies/index.html with a .bak-20260617_183550 snapshot preserved (18:35). _shared/tooltip.js refreshed (15:25). App-status watchdog went RED on strategy_bar/strategy_registry at 21:31 -- the two derived views still lack a scheduled refresh, so the monitor false-alarms after a master change. Wiring that schedule + monitor-reclass is the remaining open step (see today_focus).",
      "project": "strat_lib"
    },
    {
      "id": "vcet-v012-expansion-gate-timing-study-2026-06-17",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VCET v0.1.2 expansion-gate TIMING study published",
      "path": "/system/VCET_V012_EXPANSION_GATE_TIMING_STUDY.html",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "param-surface-db",
      "note": "New v0.1.2 study (JSON + HTML, 08:30) examining WHEN the single-knob expansion gate fires relative to the move -- the timing dimension on top of the 0A bleed-cut. Feeds the 0E stop/target/trail cell pick and the 0B scanner-filter contract.",
      "project": "atlas"
    },
    {
      "id": "scorecard-realtime-audit-2026-06-17",
      "phase": "phase-scorecards",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Scorecard real-time audit (Grade vs plan-ingest vs cockpit basis)",
      "path": "/home/trsports/trading/dev/scorecard-realtime-audit/",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "scorecards",
      "note": "Read-only audit: /scorecards/ Grade is OUTCOME-only (gen_scorecard.py:552); the plan IS ingested (stocks_in_play_daily) and partly compared but unscored, if_then inert; mood capture exists (save_journal.php); positions_snapshot.json ~30s = cockpit basis. 4 deliverables in the dev tree; first 3 changes defined. Read-only -- no live scorecard or HUD edits.",
      "project": "atlas"
    },
    {
      "id": "focus-readiness-pages-2026-06-17",
      "phase": "design",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "New /system/ focus3-plan + readiness pages published",
      "path": "/system/readiness.html",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "param-surface-db",
      "note": "Net-new system surfaces: system/focus.json (16:58), system/focus3-plan.html (19:21), system/readiness.html (19:44). Focus/readiness layer for the daily-stop / pre-trade lane. Scaffolding -- not yet wired into the hub or scheduled.",
      "project": "atlas"
    },
    {
      "id": "param-surface-monthly-regen-2026-06-17",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "VWAP-bands param-surface monthly report regenerated (daily)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/index.html",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "param-surface-db",
      "note": "Daily regen 2026-06-17 (16:37-16:38): per-symbol + combined param-surface/heatmap PNGs (CROX/QBTS/BTDR/AMZN/ALOY/MSFT) and full ToD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion. james-monthly-reports june-2026 + vwap-bands + param-surface-db indexes rebuilt.",
      "project": "atlas"
    },
    {
      "id": "scorecards-daily-2026-06-17",
      "phase": "phase-scorecards",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Daily scorecard + you-vs-you regen for 6/17",
      "path": "/scorecards/trade_scorecard_2026-06-17.html",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "scorecards",
      "note": "Daily scorecard pipeline ran clean: trade_scorecard_2026-06-17.html (16:31) + manifest.json + you-vs-you.html (16:39) regenerated.",
      "project": "multi_trader"
    },
    {
      "id": "multi-tf-stack-refresh-2026-06-17",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Multi-TF Index Stack v4 + supporting report scripts ran/refreshed for 6/17",
      "path": "/z/setup_daily_reporting/reports/gen_multi_tf_stack.py",
      "completed_at": "2026-06-17",
      "created_at": "2026-06-17",
      "property": "mgp",
      "note": "Multi-TF stack v4 ran clean for 2026-06-17: per-symbol 1D caches built pre-market (04:08) + 5m source caches refreshed through the session (08:23-10:55) for SPY/QQQ/IWM/SMH/VIX/UVXY; multi_tf_stack_daemon.py (16:54) + gen_multi_tf_stack.py (19:56) iterated; chiphazzard_v4.py stage classifier (16:05), theme_chrome.py (19:51), econ_calendar.py (06:55), start_multi_tf.cmd launcher touched; /system/multi-tf-stack/ page refreshed (18:24).",
      "project": "atlas"
    },
    {
      "id": "first-entry-analysis-2026-06-18",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "First-entry timing analysis (account 14859, Apr-Jun)",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/james-domino/first-entry-analysis/index.html",
      "property": "monthly_reports",
      "completed_at": "2026-06-18",
      "created_at": "2026-06-18",
      "note": "gen_first_entry_analysis.py pulls _trades_14859_apr_jun + per-symbol minute cache and builds a first-entry-of-day timing study page. Feeds the selectivity thesis (James over-trades; first entry quality drives the day).",
      "project": "atlas"
    },
    {
      "id": "top5-bands-generator-2026-06-18",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Top-5 VWAP-bands generator (gen_top5_bands.py)",
      "path": "/z/setup_daily_reporting/reports/gen_top5_bands.py",
      "property": "param_surface",
      "completed_at": "2026-06-18",
      "created_at": "2026-06-18",
      "note": "New generator emitting /system/top5-bands.json -- the day's top-5 symbols with VWAP-band context for the morning surface. Refreshes daily.",
      "project": "strat_lib"
    },
    {
      "id": "session-verdict-engine-2026-06-19",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "Session-verdict engine (session_verdict.py + simulate harness)",
      "path": "/z/setup_daily_reporting/reports/session_verdict.py",
      "property": "pm_system",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "New session_verdict.py + test_verdict_simulate.py -- end-of-session verdict logic with a simulate harness for offline testing. Pairs with the readiness/focus3 surfaces.",
      "project": "atlas"
    },
    {
      "id": "multi-tf-stack-daemon-2026-06-19",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "Multi-TF stack daemon + /system/multi-tf-stack page",
      "path": "/z/setup_daily_reporting/reports/multi_tf_stack_daemon.py",
      "property": "pm_system",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "multi_tf_stack_daemon.py + start_multi_tf.cmd launcher + gen_multi_tf_stack.py refactor, publishing the /system/multi-tf-stack/ surface. Moves the multi-TF stack from on-demand to a running daemon.",
      "project": "atlas"
    },
    {
      "id": "system-readiness-page-2026-06-19",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "/system/readiness page built from 6/17 scaffold",
      "path": "/z/public_html/system/readiness.html",
      "property": "pm_system",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "readiness.html filled in off the 6/17 scaffold -- a pre-trade readiness surface in the /system/ hub.",
      "project": "atlas"
    },
    {
      "id": "focus3-status-page-2026-06-19",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "/system/focus3-status page",
      "path": "/z/public_html/system/focus3-status.html",
      "property": "live_classifier",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "Focus-3 status surface (trend+backside, long+short) -- the live status view on top of the 6/19 Focus-3 backtest selectivity finding.",
      "project": "classifier"
    },
    {
      "id": "methodology-page-2026-06-19",
      "phase": "phase-foundation",
      "status": "completed",
      "owner": "claude",
      "title": "/system/methodology page",
      "path": "/z/public_html/system/methodology.html",
      "property": "foundation",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "New methodology page documenting the system's approach in the /system/ hub.",
      "project": "atlas"
    },
    {
      "id": "vwap-state-engine-v4-2026-06-19",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VWAP state engine + chip-hazard classifier v4",
      "path": "/z/setup_daily_reporting/reports/chiphazzard_v4.py",
      "property": "live_classifier",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "dale_vwap_state.py update + chiphazzard_v4.py (chase/chip-hazard classifier v4). Continues the VWAP-band-entry / chase-rule line of work toward a sized entry contract.",
      "project": "classifier"
    },
    {
      "id": "vwap-research-hub-2026-06-19",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "title": "VWAP research hub index (/gt_dev/db_reports/vwap-research)",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "property": "monthly_reports",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "New vwap-research hub index gathering the VWAP-band / param-surface research under one landing page; monthly-report index + june-2026 + vwap-bands sub-indexes regenerated alongside.",
      "project": "atlas"
    },
    {
      "id": "you-vs-you-scorecard-2026-06-19",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "/scorecards/you-vs-you view",
      "path": "/z/public_html/scorecards/you-vs-you.html",
      "property": "scorecards",
      "completed_at": "2026-06-19",
      "created_at": "2026-06-19",
      "note": "You-vs-you scorecard comparison view published alongside the daily 6/19 scorecard + manifest refresh.",
      "project": "multi_trader"
    },
    {
      "id": "price-provider-abstraction-layer",
      "phase": "phase-0",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Build price_provider.py abstraction layer \u2014 kill 17 direct kti subprocess calls so future swap (TradeStation/Polygon/etc) is 1-2 hours instead of 10-15",
      "path": "/z/setup_daily_reporting/reports/price_provider.py",
      "created_at": "2026-06-20",
      "note": "AUDIT 2026-06-19: 17+ files in setup_daily_reporting/reports call kti directly via subprocess.run(['kti', 'price', ...]) \u2014 chiphazzard_v4, mtf_trend, dale_vwap_state, breadth, stocks_in_play, morning_tickers_v2, precompute_chart_data, gen_reports, alpha_trend, build_best_*_examples, playbook_helpers, gen_me_vs_strategy_charts + 5 more. No abstraction layer exists today.\n\nSCOPE (~3-4 hours):\n  1. Build price_provider.py with fetch_daily / fetch_minute / fetch_five_min interface.\n  2. First implementation routes to kti (no behavior change).\n  3. Replace each subprocess.run(['kti', ...]) call site with price_provider.fetch_*().\n  4. Add PRICE_PROVIDER env var so future swaps are flag-flips.\n  5. Verify cockpit + MGP + scorecards still produce identical output.\n\nPAYOFF: When TradeStation API / Polygon.io / Alpaca / etc swap considered later, it becomes a 1-2 hour test instead of a 10-15 hour rewrite. Also makes mocking/testing price data trivial (currently test_verdict_simulate.py uses hand-built fake breadth dicts).\n\nNOT URGENT: kti works fine today. This is preventive technical debt cleanup that unlocks easier provider swapping in the future. Best done on a quiet afternoon.",
      "project": "atlas"
    },
    {
      "id": "ams-seed-users-from-main",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "user",
      "priority": "P0",
      "title": "AMS: seed main.users with trader rows so identity_fn returns non-zero",
      "path": "/home/trsports/trading/dev/ams/sql/migrations/",
      "created_at": "2026-06-20",
      "note": "Per AMS Phase 0 status doc: p0-identity-fn shipped + 5/5 smoke tests pass, but resolve_identity() returns 0 rows because no users in main.users have usernames matching trader short_names (james/nolan/preston/brian/thomas/jimbo). This is THE pain blocking user onboarding. Seed users (james + the 4 internal traders + jimbo) into main.users with username matching their trader short_name. Path B of identity resolution activates immediately. ~5 INSERT statements; could be a single 003_seed_users.sql migration.",
      "property": "foundation",
      "project": "ams",
      "sprint": 1
    },
    {
      "id": "ams-migration-002-remembered-devices",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "title": "AMS: create main.remembered_devices table (migration 002) \u2014 Phase 1 hard dependency",
      "path": "/home/trsports/trading/dev/ams/sql/migrations/002_remembered_devices.sql",
      "created_at": "2026-06-20",
      "note": "Per AMS Phase 0 status doc 'Known gaps': table required for 2FA 'remember-device' feature. Must ship before Phase 1 sign-off can happen. Idempotent CREATE TABLE IF NOT EXISTS, targets *_dev first per AMS rule 2.",
      "property": "foundation",
      "project": "ams",
      "sprint": 1
    },
    {
      "id": "ams-decisions-d1-through-d7",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "user",
      "priority": "P0",
      "title": "AMS: resolve 7 open decisions D1-D7 (canonical domain \u00b7 auth stack \u00b7 DB strategy \u00b7 sharing \u00b7 risk_hud \u00b7 2FA \u00b7 daily-stop)",
      "path": "/home/trsports/trsports_gd/html/system/AMS-INTEGRATION-PLAN-v1.md",
      "created_at": "2026-06-20",
      "note": "Per AMS README \u00a7'Open decisions': D1 canonical domain (rec: smb.a1prospect.com), D2 auth stack (rec: stay PHP), D3 DB strategy (rec: build in _dev, promote together), D4 sharing default (rec: opt-in via willing_to_share, bulk-enable for internal), D5 risk_hud fate (rec: archive \u2014 accountability-v2 subsumes), D6 mandatory 2FA roles (rec: required admin+enterprise_admin; trader optional+remembered-device), D7 Daily Stop tool (rec: keep at /daily-stop/, just wrap in auth). Claude has recommendations. James to sign off \u2014 bulk-accept if recs are good, or note exceptions.",
      "property": "foundation",
      "project": "ams",
      "sprint": 1
    },
    {
      "id": "ams-phase3-status-doc",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS: write PHASE-3-STATUS.md mirroring PHASE-0-STATUS.md (Phase 3a baseline shipped, what's left)",
      "path": "/home/trsports/trading/dev/ams/docs/PHASE-3-STATUS.md",
      "created_at": "2026-06-20",
      "note": "Per git log, AMS shipped 'Phase 3a baseline: scope + privacy + cockpit port + admin tooling + auth polish'. Plus: own-copy MGP, own-copy scorecards, live per-trader scorecard (/scorecards/live.php), shell consistency across 22 pages, HTTPS enforcement, dev-profile DB path. Atlas (PM cockpit) had zero visibility. Write Phase 3 status doc to mirror PHASE-0-STATUS.md format. Reading it should answer: what shipped, what's left in Phase 3, when does Phase 4 start.",
      "property": "foundation",
      "project": "ams",
      "sprint": 1
    },
    {
      "id": "ams-apply-migrations-021-022-to-dev",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "AMS: apply pending accountability migrations 021 (trade_action_requests) + 022 (hud_usage_events) to dev DB",
      "path": "/home/trsports/trading/dev/a1p_journal/accountability-v2/sql/",
      "created_at": "2026-06-20",
      "note": "Per AMS Phase 0 status 'Known gaps': two tables expected per inventory but not in trsports_tradingjournal_dev. Migration 021 (trade_action_requests) blocks the hud-orderlab close/stop intent feature. Migration 022 (hud_usage_events) is analytics-only. Non-blocking for Phase 1 but cleanup needed during Phase 4A cockpit wiring.",
      "property": "accountability_hud",
      "project": "ams"
    },
    {
      "id": "ams-cross-link-todos-cockpit",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS: add AMS-side cockpit link back to /todos/ + sync ams/todos/ with trsportsacademy.com/todos/",
      "path": "/home/trsports/trading/dev/ams/todos/",
      "created_at": "2026-06-20",
      "note": "AMS has its own todos/ directory (per dir listing). To prevent two-source drift: either (a) /gt_dev/ams/todos/ becomes a symlink to /trsports_gd/html/todos/, or (b) one is canonical and the other is read-through. Either way, only ONE place stores todos.json. After Atlas added project codenames + per-project Gantt, AMS users need to see them too.",
      "property": "pm_system",
      "project": "ams"
    },
    {
      "id": "exec-trigger-engine-mvp",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Exec: Trigger Engine MVP \u2014 poll classify/*.json, detect state changes, emit Alert records",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "Stage 1 of Execution Pipeline. Reads /scorecards/classify/<SYMBOL>.json every minute during RTH. When a symbol transitions from NOT_APPLICABLE \u2192 TRIGGERED with gates_passed=N/N on a hot playbook, write an Alert record to alerts_tradingview.execution_triggers (new table). No alerts sent yet \u2014 just data in DB. Depends on classifier-refresh-cron (Classifier project) shipping first.",
      "property": "live_classifier",
      "project": "exec"
    },
    {
      "id": "exec-alert-channels-sms-email-push",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Exec: Alert Channels \u2014 Twilio SMS + existing mailer + in-HUD WebSocket push",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "Stage 2. When Trigger Engine writes an Alert, fan out: (a) Twilio SMS to James/internal traders, (b) email via existing _api/mailer.php, (c) WebSocket push to Accountability HUD chat sidebar. Per-user mute schedules. Depends on Trigger Engine.",
      "property": "live_classifier",
      "project": "exec"
    },
    {
      "id": "exec-risk-gate",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Exec: Risk Gate \u2014 pre-trade check (daily_stop remaining \u00b7 ticket count \u00b7 sizing) BEFORE order",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "Stage 3. Sits between Trigger and broker. Reads from /scorecards/api/user_settings.php (Daily Stop) + accountability-v2 /api/health.php (current capital usage). Refuses to fire order if any limit breached. Logs rejection reason. Depends on Trigger Engine + Alert Channels.",
      "property": "daily_stop",
      "project": "exec"
    },
    {
      "id": "exec-broker-adapter-gr8trade",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Exec: Gr8Trade adapter \u2014 translate Alert \u2192 Gr8Trade order via Kore bridge",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "Stage 4. First broker adapter. Pattern: gr8.send_order(symbol, side, qty, type, stop, target) \u2192 uses existing Kore tick bridge from accountability-v2. Forward-test only initially (no real orders). Conservative caps: max 1 share, max 1 order/day, single-symbol whitelist. Depends on Risk Gate.",
      "property": "accountability_hud",
      "project": "exec"
    },
    {
      "id": "exec-broker-adapter-ibkr",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Exec: IBKR adapter (Phase 2 of execution rollout \u2014 after Gr8Trade in forward test)",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "IBKR has the most stable broker API of the three. Adapter pattern same shape as Gr8Trade. Depends on Gr8Trade adapter being live and stable for 2+ weeks.",
      "property": "foundation",
      "project": "exec"
    },
    {
      "id": "exec-broker-adapter-etrade",
      "phase": "phase-strategy-rules",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Exec: eTrade adapter (Phase 3 of execution rollout \u2014 last broker, hardest API)",
      "path": "tbd-new-service",
      "created_at": "2026-06-20",
      "note": "eTrade has the worst-documented API of the three. Adapter goes last after IBKR proves the pattern.",
      "property": "foundation",
      "project": "exec"
    },
    {
      "id": "backup-mirror-scheduled-tasks-to-jimdo",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Backup: mirror scheduled-task SKILL files from domino-1 to jimdo via scp",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\",
      "created_at": "2026-06-20",
      "note": "Step 1 of warm-backup architecture. scp -r ~/.claude/scheduled-tasks/* jimdo:/c/Users/jimdo/.claude/scheduled-tasks/. Plus a sync script that runs nightly to keep them in lockstep. Now possible since `ssh jimdo` is wired (verified 2026-05-13).",
      "property": "pm_system",
      "project": "backup",
      "sprint": 1
    },
    {
      "id": "backup-jimdo-task-delay-and-skip",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Backup: add 5-min delay + 'check if primary already wrote output, exit if so' to JIMDO's SKILL Step 0",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\nightly-project-manager\\SKILL.md",
      "created_at": "2026-06-20",
      "note": "Step 2 of warm-backup. JIMDO's nightly-project-manager fires at 23:09 ET (5 min after domino-1's 23:04). Step 0 reads meta.last_nightly_report_date from todos.json \u2014 if it == today's date, exit silently. Otherwise proceed as if it were the primary. Same pattern for mgp-daily / mgp-midday / mgp-recap.",
      "property": "pm_system",
      "project": "backup"
    },
    {
      "id": "backup-heartbeat-domino-to-system",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Backup: domino-1 writes /system/heartbeat.json every 5 min; jimdo reads it for liveness",
      "path": "/z/public_html/system/heartbeat.json",
      "created_at": "2026-06-20",
      "note": "Step 3. Tiny Python script on domino-1: every 5 min POSTs {timestamp, host, status} to /system/heartbeat.json. JIMDO's pre-fire check reads it \u2014 if domino-1 heartbeat is >10 min stale during a scheduled-task window, JIMDO auto-promotes for that fire (skips the 5-min delay).",
      "property": "pm_system",
      "project": "backup"
    },
    {
      "id": "backup-ssh-jimdo-liveness-check",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Backup: domino-1 nightly task runs `ssh jimdo 'tailscale status'` to confirm peer is alive + key still works",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\nightly-project-manager\\SKILL.md",
      "created_at": "2026-06-20",
      "note": "Step 4 \u2014 defense in depth. Each night, primary verifies it can still SSH to jimdo. If `ssh jimdo` fails, log a SILENT-FAIL-NEXT-TIME warning to notify.log and the cockpit freshness panel goes red on backup-peer-stale. Catches the case where Tailscale signs out on JIMDO and we don't notice for weeks.",
      "property": "pm_system",
      "project": "backup"
    },
    {
      "id": "backup-james-trader-machine-onboarding",
      "phase": "phase-project-mgmt",
      "status": "pending",
      "owner": "user",
      "priority": "P2",
      "title": "Backup: add James's new trading machine to Tailscale + ~/.ssh/config + machine roster + freshness probe",
      "path": "C:\\Users\\jtdom\\.ssh\\config",
      "created_at": "2026-06-20",
      "note": "Per machine roster: James's new trading machine is planned but not yet onboarded. When it arrives: install Tailscale, sign in to same account, add SSH alias `james-trader` to ~/.ssh/config, document in reference_machine_roster.md, optionally add as third-tier backup runner. This is the machine that will run the Exec broker adapters.",
      "property": "foundation",
      "project": "backup"
    },
    {
      "id": "scrum-sprint-cadence-layer-2026-06-20",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Scrum/sprint cadence layer \u2014 /system/sprint-plan.html + /system/scrum-cadence.html",
      "path": "/z/public_html/system/sprint-plan.html",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-20",
      "note": "SHIPPED 2026-06-20 (Sat). Stood up a scrum-master view over the 361-task backlog: Sprint 1 ACTIVE (6/20\u21926/27, theme = Foundation hardening + AMS unblock) with live tickets loaded from todos.json (sprint field), Sprints 2-6 queued (classifier production, Exec trigger engine/alert channels/risk gate, Gr8Trade adapter forward-test, broker adapters), a 'Today \u00b7 what shipped' changelog, and scrum-cadence.html. Backlog is now a plan, not a pile.",
      "property": "todo",
      "project": "atlas"
    },
    {
      "id": "todos-project-tagging-registry-2026-06-20",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Tag all 361 tasks with project + projects registry + per-project Gantt + /system/ project cards",
      "path": "/z/public_html/todos/gantt.html",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-20",
      "note": "SHIPPED 2026-06-20 (Sat). Every task tagged with a `project` field; top-level `projects` registry added to todos.json; Projects section (9 data-driven cards) added to /system/ hub, each click-through to a per-project Gantt; per-project Gantt filter /todos/gantt.html?project=<id> with switcher chip row + project-aware title/subtitle. Codename registry (9 projects + 5 AI roles) locked in memory.",
      "property": "todo",
      "project": "atlas"
    },
    {
      "id": "ams-a2-inventory-phase3a-baseline-2026-06-20",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "AMS a2 inventory \u2014 discovered Phase 3a baseline (not stalled) + 6 catch-up tasks filed",
      "path": "/home/trsports/trading/dev/ams/",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-20",
      "note": "SHIPPED 2026-06-20 (Sat). Inventoried AMS on a2 via git: AMS is at Phase 3a baseline (scope + privacy + cockpit port + admin tooling + auth polish shipped; MGP + Scorecards migrated from symlinks to AMS-owned copies) \u2014 not stalled as the plan assumed. Filed 6 catch-up tasks incl. the headline onboarding blocker (seed main.users), migration 002 remembered_devices, decisions D1-D7, Phase-3 status doc, apply migrations 021/022 to dev. Exec (6) + Backup (5) roadmap tasks also scoped.",
      "property": "foundation",
      "project": "ams"
    },
    {
      "id": "focus3-board-republish-2026-06-20",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Focus-3 board republished \u2014 plan + status + morning readiness",
      "path": "/z/public_html/system/focus3-plan.html",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-20",
      "note": "Republished 2026-06-20 (Sat). Focus-3 three-regime plan (Backside Long / Trend Trade / TD VWAP) with ALL-gates-or-no-trade SVG band maps + rotation\u2192trend breakout rule, focus3-status.html, and the Morning Readiness board (live regime read + breakout alerts). _shared/jd_nav.js refreshed.",
      "property": "strategy_library",
      "project": "strat_lib"
    },
    {
      "id": "param-surface-scorecards-refresh-2026-06-20",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Param Surface + monthly reports + scorecards weekend refresh",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-20",
      "note": "Regenerated 2026-06-20 (Sat). VWAP-bands Param Surface refreshed: per-symbol surface+heatmap PNGs (CROX/QBTS/MSFT/BTDR/AMZN + combined), full TOD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion, param-surface-db index. james-monthly-reports june-2026 + vwap-bands index rebuilt. Scorecards: trade_scorecard_2026-06-20.html + manifest + you-vs-you.html regenerated. gen_multi_tf_stack.py touched.",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "sip-spec-composite-scorer",
      "title": "SIP composite scoring formula spec (weights, refresh cadence, gating rules)",
      "priority": "P0",
      "status": "completed",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "completed_at": "2026-06-21",
      "note": "Spec realized as working code 2026-06-21: sip_scorer.py implements score = 40*catalyst + 30*rrg + 20*mover + 10*liquidity, each sub-score normalized [0..1] then weighted, final clipped [0,100]. Allowlist baseline is the FLOOR (allowlist wins if higher) per James's 2026-06-20 lock. Pure-Python dataclass inputs (CatalystInput/RrgInput/MoverInput). sip_scorer_extension.py adds the extended dimensions."
    },
    {
      "id": "sip-catalyst-feed",
      "title": "Catalyst feed: Vital Dawn + Finnhub news + SEC 8-K + Benzinga RSS into per-symbol catalyst_score",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Vital Dawn already parsed by MGP \u2014 reuse. Finnhub free tier 60/min. SEC EDGAR free. Output: per-symbol score 0-1 + reason string."
    },
    {
      "id": "sip-rrg-engine",
      "title": "RRG engine: RS-Ratio + RS-Momentum from KITE daily bars \u00b7 4-quadrant classifier",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Daily refresh pre-market + 30-min during RTH. Leading (1.0) > Improving (0.7) > Weakening (0.3) > Lagging (0.0). Benchmark: SPY (override per sector if needed)."
    },
    {
      "id": "sip-mover-detector",
      "title": "Daily mover detector: pre-market gappers + top dollar-volume + ATR expansion + always-active allowlist",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Always-active: AMZN/TSLA/NVDA/AAPL/META/COIN/PLTR/MSTR/GME/AMC + James's discretionary additions. Liquidity gate: dollar_volume > $20M today."
    },
    {
      "id": "sip-publish-watchlist",
      "title": "Publish ranked watchlist.json + /sip/index.html (top 15-20 \u00b7 refreshes pre-market + every 30m RTH)",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "watchlist.json consumed by Trigger Listener (Exec). /sip/index.html shows ranked names with score breakdown, click \u2192 SymbolView modal."
    },
    {
      "id": "sip-integrate-with-trigger-listener",
      "title": "Integrate SIP into Trigger Listener: gate alerts on watchlist \u00b7 enrich payload with score + catalyst + RRG",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Only fire alert if symbol in current SIP top-15. Alert payload includes: SIP score, catalyst reason, RRG quadrant, mover stats."
    },
    {
      "id": "sip-historical-validation",
      "title": "Historical validation: backtest 30d MGP-stocks vs SIP-top-20 \u2014 confirm SIP outperforms (more follow-through, fewer chops)",
      "priority": "P2",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Use existing scorecards to measure follow-through. Hypothesis: SIP-gated alerts have higher avg-R and lower chop rate vs MGP-only."
    },
    {
      "id": "sip-briefing-com-integration",
      "title": "Briefing.com integration: Story Stocks + In Play + Pre/After-Hours Movers as primary catalyst source",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "Sections: https://www.briefing.com/after-hours-report (Story Stocks, In Play). Free + public. Scrape with cache + retry. James authorizes for personal use per Briefing ToS. Fallback chain: Briefing -> Finnhub -> SEC -> Vital Dawn (last)."
    },
    {
      "id": "sip-allowlist-finalize",
      "title": "Lock the SIP always-active allowlist (24 names: mega-cap tech + semis + crypto + AI + index + high-beta retail)",
      "priority": "P1",
      "status": "completed",
      "owner": "user",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "LOCKED 2026-06-20 with 25 names including SPCX (SpaceX direct IPO 2026). Saved to /sip/allowlist.json. Categories: mega_cap_tech (8) + semis (4) + crypto (3) + ai_data_space (4) + index_gateways (2) + high_beta_retail (4). Baseline score 20-25 each. Composite scorer reads this file.",
      "completed_at": "2026-06-20"
    },
    {
      "id": "sip-vital-dawn-deprecation-evaluation",
      "title": "Evaluate Vital Dawn deprecation impact (if James cancels SMB-bundled feed): swap for Briefing.com as primary",
      "priority": "P2",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-20",
      "note": "If Vital Dawn goes away, MGP narrative loses one source but Briefing.com Story Stocks + In Play covers ~80% of same content. Document the swap path: parser changes, MGP section template updates, deprecation timeline. Decision: keep VD if free; cancel if paid extra."
    },
    {
      "id": "pinepb-codify-rules-three-plays",
      "title": "Codify VWAP Pull-In + Trend Trade + Backside Scalp rules in Pine-compatible terms",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Memory says rules conceptually exist but never coded. VWAP Pull-In: gap + MA stack + VWAP state + entry grades (Cons/Mod/Agg). Backside Scalp: gap-up failure into PDH/2-day VWAP/DMA. Trend Trade: trend continuation. Each needs explicit entry trigger, stop, target, management rules."
    },
    {
      "id": "pinepb-write-pine-v5-script",
      "title": "Write Pine v5 indicator combining all 3 playbooks with toggle inputs",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Single indicator, 3 toggle groups (one per play). Each play has its own entry/stop/target plot. Conditions are valid Pine v5 (TradingView Editor compiles cleanly). Inspired by existing FashLate Pine architecture."
    },
    {
      "id": "pinepb-alert-webhook-template",
      "title": "Alert message JSON template matching alerts_tradingview.indicator_alerts schema",
      "priority": "P0",
      "status": "completed",
      "owner": "claude",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Shipped 2026-06-21: three alert-payload JSON templates with TradingView placeholders ({{ticker}}/{{close}}/{{timenow}}/{{strategy.order.action}}) + plot() pulls for vwap/sigma/bands/regime/slope/stop/grade \u2014 pine_vwap_pull_in_template.json (playbook_type=vwap_pullback_continuation), pine_fl2_fash_late_template.json, pine_mtf_vp_template.json. Each carries the new playbook_type/setup_grade/band_slope fields the Trigger Listener consumes. Pine v5 indicator (pinepb-write-pine-v5-script) is the next P0.",
      "completed_at": "2026-06-21"
    },
    {
      "id": "pinepb-tradingview-setup-guide",
      "title": "TradingView setup guide for James \u2014 paste Pine script, configure alerts, test",
      "priority": "P0",
      "status": "pending",
      "owner": "user",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Click-by-click instructions. Save indicator to user's library. Add to chart. Configure alert with webhook URL. Test fire. ~15 min one-time."
    },
    {
      "id": "pinepb-backtest-validation",
      "title": "Backtest each play on historical data \u2014 validate rules match what James intuits",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Use 60d of bars on 5 known James trades per play. Each play's signal should fire on the historical bars where the play actually paid. If misses or false-fires, iterate rules."
    },
    {
      "id": "pinepb-grade-classifier",
      "title": "Per-play grade classifier (A/B/C) based on context (MA stack, RVOL, VWAP state)",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "Each play emits setup_grade in the alert payload. Trigger Listener uses grade to determine sizing (A=full, B=half, C=skip per SMB framework). Grade derived from runtime context, not hardcoded."
    },
    {
      "id": "pinepb-deploy-to-james-tradingview",
      "title": "James pastes finalized Pine script into TradingView + configures the 3 alerts",
      "priority": "P0",
      "status": "pending",
      "owner": "user",
      "project": "pinepb",
      "created_at": "2026-06-21",
      "note": "After Atlas delivers the script + setup guide, James spends ~15 min configuring TradingView. Each play gets its own alert (3 alerts total) so they can be toggled independently."
    },
    {
      "id": "sip-runtime-delivery-layer-2026-06-21",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P0",
      "title": "SIP runtime: MVP daily watchlist generator + MGP catalyst integration + Telegram delivery + Trigger Listener",
      "path": "/z/setup_daily_reporting/reports/run_sip_daily.py",
      "created_at": "2026-06-21",
      "completed_at": "2026-06-21",
      "note": "Built the SIP runtime layer 2026-06-21 (Sun). run_sip_daily.py/.sh = MVP Path D daily generator: reads alerts_tradingview.stocks_in_play_daily, builds SipInput (allowlist baseline + grade->catalyst proxy), runs the v2 scorer, writes /z/public_html/sip/watchlist.json. mgp_sip_integration.py wires Vital Dawn -> CatalystInput (source classified earnings/fda/guidance/analyst/m&a) for paste into gen_mgp.py. telegram_sender.py + telegram.log = new Telegram delivery channel. trigger_listener.py + run_listener.sh = Exec Trigger Listener service scaffold. test_a2_live.py + run_all_eod.bat/.vbs + setup_eod_silent_task.ps1/.cmd = silent EOD task automation. NOTE: /sip/index.html UI + Finnhub/SEC/Benzinga catalyst sources + RRG/mover dimensions still pending (later sprint).",
      "property": "mgp",
      "project": "sip"
    },
    {
      "id": "param-surface-scorecards-refresh-2026-06-21",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "Param Surface + monthly reports + scorecards weekend refresh",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "created_at": "2026-06-21",
      "completed_at": "2026-06-21",
      "note": "Regenerated 2026-06-21 (Sun). VWAP-bands Param Surface refreshed: per-symbol surface+heatmap PNGs (CROX/QBTS/MSFT/BTDR/AMZN + combined), full TOD-bucket surface/heatmap set (open/early/mid-am/midday/mid-pm/late) + tod_comparison + tod_sd_expansion, param-surface-db index. june-2026 + vwap-bands indices rebuilt. Scorecards: trade_scorecard_2026-06-21.html + manifest + you-vs-you.html regenerated.",
      "property": "param_surface",
      "project": "strat_lib"
    },
    {
      "id": "ui-grade-tooltips-2026-06-22",
      "title": "Page-wide tooltip system \u2014 every term, score, regime, threshold gets a What/Why/How/Use/Source hover",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "mtfvp",
      "created_at": "2026-06-22",
      "note": "Originally scoped to grades. James 2026-06-22 confirmed: tooltips on EVERYTHING that needs explanation. Each tooltip follows the template WHAT/WHY/HOW/USE/SOURCE (1-2 sentences each). Targets: every setup name (LYNCH/VCET regimes/VWAP Pull-In/Trend/Backside), every score (composite/grade/confidence), every metric (gap %/RVOL/ATR/distance to breakout), every regime label (TREND_UP/ROTATION), every threshold value, every formula. Implementation: single tooltip-system.js that hover-targets data-tt attributes; tooltip content lives in tooltips.json keyed by (term_id) -> {what, why, how, use, source}. New setups add their entries to that JSON; viewer renders automatically with consistent styling. Surfaces affected: /sip/ (primary), /mgp/ brief (mirror), /scorecards/, /system/multi-tf-stack/ sentiment grade panel (contributors breakdown)."
    },
    {
      "id": "mtfvp-coc-event-timeline-2026-06-22",
      "title": "Change-of-Character panel \u2014 timestamp + duration + severity-trend per row",
      "priority": "P1",
      "status": "in_progress",
      "owner": "claude",
      "project": "mtfvp",
      "created_at": "2026-06-22",
      "note": "James spotted that the /system/multi-tf-stack/ CoC summary lists symbols (IWM daily 2-2 -> 15m=1-3, SMH daily 2-2 -> 5m=3-1) without telling you WHEN the signature first appeared or how long it has been active. Data already exists in alerts/<date>.jsonl with ts_local per transition. Phase A (shipping now, 2026-06-22): pull the earliest matching event from today's alert log for each CoC row, inject 'First seen HH:MM ET / active Nm' line. Phase B (later): severity-trend arrow (rising/stable/fading) by counting recent transitions in the same kind. Phase C (later): full collapsible event timeline + sparkline of grade vs time pinned beneath the panel."
    },
    {
      "id": "mgp-sip-read-from-db-2026-06-22",
      "title": "MGP brief reads stocks_in_play_daily from DB (no kti recompute)",
      "priority": "P0",
      "status": "in_progress",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Today's bug: gen_mgp.py:_build_and_render_sip() does live kti calls every rebuild. When kti rate-limits, the brief shows 'bar fetch failed' while the SIP page shows real picks (different sources of truth). Fix: read alerts_tradingview.stocks_in_play_daily for today's date, render directly from the persisted rows. Same-day win \u2014 kills the MGP/SIP divergence James spotted."
    },
    {
      "id": "sip-master-picker-unification-2026-06-22",
      "title": "SIP becomes master picker \u2014 multi-source scanner union writes stocks_in_play_daily",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Architectural flip: today MGP picks \u2192 SIP ranks. New: SIP picks (multi-source scanner union) \u2192 MGP/viewer/listener all CONSUME from stocks_in_play_daily. Single source of truth. Scanners: catalyst (VK + Briefing + KTI gappers + RSS), LYNCH continuation breakout, VCET regime, RRG-leading, momentum, mega-cap floor. Each scanner tags every candidate with (setup_name, source, confidence, reason). Dedup by ticker. Composite rank = sum of weighted confidences + consensus bonus when >=2 scanners flag same name. Top 5 actionable + 5 watching written to DB."
    },
    {
      "id": "sip-two-pass-schedule-2026-06-22",
      "title": "Schedule: SIP Pass 1 (04:15 ET) + SIP Pass 2 (05:45 ET) + MGP (06:10 ET)",
      "priority": "P0",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "New cadence: 04:15 SIP Pass 1 (overnight catalyst sweep \u2014 earnings, Asia/Europe, early gappers). 05:45 SIP Pass 2 (Vital Knowledge integration ~05:20 arrival, confirmed pre-market gappers, LYNCH scan, VCET classify, finalize universe). 06:10 MGP brief reads finalized DB and renders. 06:30 /sip/ viewer publish (moved earlier than current 06:40). 07:00 Trigger Listener cron (unchanged). Tasks: create 2 new Windows scheduled tasks for SIP Pass 1+2; modify existing MGP task time; update a2 cron for /sip/ runner."
    },
    {
      "id": "sip-briefing-com-feed-2026-06-22",
      "title": "Add Briefing.com after-hours as a catalyst source (free fallback to Vital Knowledge)",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Add Briefing.com after-hours feed as a parallel catalyst source alongside Vital Knowledge. Both run by default for redundancy; either can be disabled in settings.json. When both flag same ticker, it gets a 2-source consensus bonus in the SIP composite. URL provided 2026-06-22: https://www.briefing.com/after-hours-report (after-hours report \u2014 free, public). Next steps: wire up the fetcher (similar to vital_dawn_fetcher.py), add to SIP catalyst scanner."
    },
    {
      "id": "sip-continuation-breakout-picker-2026-06-22",
      "title": "LYNCH Continuation Breakout scanner module (per Google Doc spec)",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Implement Continuation Breakout picker per docs.google.com/document/d/1F0U7w1CYtlYIG97SX9bycjYgaRxqTNQCg-sLckNhfSU. First leg >=15% in 3-7 days (linear, R2>=0.70, >=70% up days, <=1 big-red day). Consolidation 2-10 days with tightening volatility, <=1 4% breakdown. Sector filter: Tech/Healthcare/ Consumer Discretionary. LYNCH score (2/L/Y/N/C/H) >=5 with L+C mandatory. Entry: breakout above range high in first hour. Stop: day low (beginner) or half entry-day range (experienced). Target: 8-20% (low-priced) or $5-50 (high-priced). 80/20 scale + kill if no move in 2 days. Run at 15:30 ET as anticipation scan for next morning + integrate into SIP Pass 2."
    },
    {
      "id": "sip-vcet-regime-categories-2026-06-22",
      "title": "VCET regime tagging on every SIP pick + 4 new viewer categories",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Run VCET classifier on every SIP candidate at pick time. Tag row with vcet_regime (ROTATION/TREND_UP/TREND_DOWN/UNKNOWN). Add 4 new categories to the /sip/ viewer: vcet_trend_long (TREND_UP + above-VWAP + dma stack), vcet_trend_short (TREND_DOWN + below-VWAP), vcet_rotation (ROTATION + near +-1sigma), vcet_continuation (TREND_UP + tight day-to-day variance, supports LYNCH). Each category maps to a James playbook. Trigger Listener can then route TradingView alerts by playbook match."
    },
    {
      "id": "mgp-long-short-thesis-2026-06-22",
      "title": "MGP brief: long thesis + short thesis + recommended bias per pick",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Per James 2026-06-22: brief should show BOTH long and short play thoughts per pick, plus a recommended bias with reason. DB schema adds: long_thesis_text, short_thesis_text, recommended_bias (LONG/SHORT/NEUTRAL), bias_reason, why_picked. SIP picker populates these at write time from per-scanner reasoning. MGP renders structured block per pick: catalyst -> long play -> short play -> recommended bias + reason -> why_picked. Single block ~10 lines per pick x 5 picks fits the brief format."
    },
    {
      "id": "sip-why-column-tooltip-2026-06-22",
      "title": "SIP viewer: surface setup_name + catalyst + reason per pick (the 'why' column)",
      "priority": "P1",
      "status": "pending",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "note": "Add a 'Why' column or hover-tooltip to each pick on /sip/ viewer. Show setup_name + catalyst + if_then + entry_hint + stop_hint + target_hint + why_picked. Data already exists in stocks_in_play_daily; viewer just needs to render it. ~30 lines in /sip/index.html. Naturally answers the question 'why is this on the list' for every pick."
    },
    {
      "id": "console-no-window-sweep-2026-06-22",
      "title": "Pass CREATE_NO_WINDOW on every kti/ssh subprocess.run across reports/",
      "priority": "P1",
      "status": "in_progress",
      "owner": "claude",
      "project": "atlas",
      "created_at": "2026-06-22",
      "note": "James 2026-06-22: scheduled scripts must run headless \u2014 no console flash. continuation_breakout.py patched. Sweep needed across all other modules that call subprocess.run([kti, ...] or [ssh, ...]): chiphazzard.py (line ~214), chiphazzard.py:_fetch_five_min (line ~393), chiphazzard_v4.py:classify_index_panel_v4 (line ~419), chiphazzard_v4.py SSH breadth fallback (line ~904), gen_multi_tf_stack_tick.py:fetch_last_prices/fetch_intraday_bars/fetch_daily_bars, run_sip_daily.py / run_sip_daily.sh (a2 side -- no Windows issue), seed_today_picks.py (already does subprocess for kti, needs flag). Pattern: add module-level CREATE_NO_WINDOW=0x08000000 + use creationflags only on win32. Reference: feedback_console_window_hiding memory. PROGRESS 2026-06-22: chiphazzard.py + chiphazzard_v4.py both patched today; remaining = gen_multi_tf_stack_tick.py + seed_today_picks.py."
    },
    {
      "id": "sip-continuation-breakout-module-2026-06-22",
      "phase": "build",
      "title": "LYNCH Continuation Breakout module built + test harness",
      "priority": "P1",
      "status": "completed",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "completed_at": "2026-06-22",
      "path": "/z/setup_daily_reporting/reports/continuation_breakout.py",
      "property": "mgp",
      "note": "Shipped 2026-06-22: continuation_breakout.py implements the LYNCH Continuation Breakout picker (first leg >=15% in 3-7 days, tightening consolidation, LYNCH 2/L/Y/N/C/H score gate). Backed by a full test harness: seed_lynch_test.py + _lynch_seed.sql seed fixtures, run_lynch_test.py executes, _lynch_test_results.json captures output. Also CREATE_NO_WINDOW-patched per the console sweep. Remaining (separate task sip-continuation-breakout-picker): wire the 15:30 ET anticipation scan + integrate into SIP Pass 2."
    },
    {
      "id": "sip-next-day-preview-builder-2026-06-22",
      "phase": "build",
      "title": "SIP next-day preview builder (v1 + v2) -> _sip_preview_2026-06-23",
      "priority": "P1",
      "status": "completed",
      "owner": "claude",
      "project": "sip",
      "created_at": "2026-06-22",
      "completed_at": "2026-06-22",
      "path": "/z/setup_daily_reporting/reports/build_sip_preview_v2.py",
      "property": "mgp",
      "note": "Shipped 2026-06-22: build_sip_preview.py (v1) then build_sip_preview_v2.py (v2 iteration), each rendering a next-session SIP preview page -- _sip_preview_2026-06-23.html and _sip_preview_v2_2026-06-23.html. Lets James eyeball tomorrow's candidate list the night before. NOTE: still a manual/preview artifact -- the live pre-market run depends on sip-two-pass-schedule."
    },
    {
      "id": "eod-brief-generator-2026-06-22",
      "phase": "build",
      "title": "EOD brief generator + catch-up automation",
      "priority": "P2",
      "status": "completed",
      "owner": "claude",
      "project": "atlas",
      "created_at": "2026-06-22",
      "completed_at": "2026-06-22",
      "path": "/z/public_html/todos/gen_eod_brief.py",
      "property": "todos",
      "note": "Shipped 2026-06-22: todos/gen_eod_brief.py builds an end-of-day brief; eod_catchup.ps1 + eod_catchup.log provide the catch-up runner so a missed EOD slot backfills on next wake. Wrote system/briefings output. Complements the silent EOD task automation landed Sunday."
    },
    {
      "id": "marketchameleon-data-integration",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "claude",
      "priority": "P3",
      "title": "MarketChameleon API integration \u2014 implied volatility / IV rank / earnings dates / unusual options activity into cockpit + MGP",
      "path": "/z/setup_daily_reporting/reports/marketchameleon_data.py",
      "created_at": "2026-06-25",
      "note": "User has a free account at marketchameleon.com but their free tier likely web-UI only. The valuable data behind login:\n  - Implied Volatility per ticker + IV Rank / IV Percentile\n  - Per-stock earnings calendar + historical earnings move %\n  - Unusual options activity (institutional flow leading indicator)\n  - Open interest / volume by strike\n\nPREREQUISITE: get an API key on the paid tier (free tier likely doesn't expose API). User decides whether the ~$60-100/mo cost justifies it.\n\nIF integrating (when API key in env var):\n  1. Build marketchameleon_data.py \u2014 fetch IV/earnings/UOA for cockpit tickers (SPY/QQQ/IWM/SMH + active scorecard symbols)\n  2. Add IV-aware sentiment contributor (high IV + price up = risk-on confirmed; high IV + price down = real fear)\n  3. Earnings calendar overlay on stocks_in_play picks (flag tickers reporting this week)\n  4. Tooltip-wire each new metric\n\nALTERNATIVE: Polygon.io $79/mo Developer tier also has IV + extended options data via a cleaner API. Compare before committing to MC.\n\nBLOCKED ON: user decision (a) whether to pay for API tier, (b) which provider (MC vs Polygon)."
    },
    {
      "id": "econ-calendar-ism-bea-sentiment",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "econ_calendar.py \u2014 add ISM Manuf/Services + BEA GDP/PCE + Conference Board + UMich Sentiment + Fed-speak (compute-by-cadence pattern)",
      "path": "/z/setup_daily_reporting/reports/econ_calendar.py",
      "created_at": "2026-06-25",
      "note": "Same compute-schedule pattern as existing fetch_jobless_claims. No Cloudflare, no scrape fragility, no ToS exposure. Closes the biggest gaps in MGP econ calendar:\n\n  - ISM Manufacturing PMI (1st business day, 10:00 AM ET)\n  - ISM Services PMI (3rd business day, 10:00 AM ET)\n  - BEA GDP (advance/2nd/3rd estimates, ~4 weeks apart starting late Jan/Apr/Jul/Oct)\n  - BEA PCE / Personal Income (monthly, ~4 weeks lag)\n  - Conference Board Consumer Confidence (last Tuesday, 10:00 AM ET)\n  - UMich Sentiment Preliminary (2nd Friday, 10:00 AM ET)\n  - UMich Sentiment Final (4th Friday, 10:00 AM ET)\n\nPlus tooltip.js entries for each new release. Invalidate today's econ-events cache after deploy so new sources show up immediately.\n\nWhy compute vs scrape: cadence is publicly documented and fixed. Scraping ISM/Conference Board/UMich pages would break the first time they redesign. Computing is bulletproof and lets the function survive without network calls.\n\nCOMPLETED 2026-06-25: 3 new fetch functions added to econ_calendar.py (fetch_ism_events, fetch_bea_events, fetch_sentiment_events) using compute-by-cadence pattern. 6 new TT keys in tooltip.js. Today's cache invalidated + cockpit rebuilt \u2192 6 new events surfaced (GDP today, PCE+UMich tomorrow, Confidence next Tue, ISM Manuf next Wed, ISM Services next Fri). Auto-propagates to all 5 consumers: gen_mgp.py, gen_mgp_v2.py, gen_mgp_ams.py, gen_multi_tf_stack.py, morning_data.py.",
      "completed_at": "2026-06-25"
    },
    {
      "id": "j3p-v22-webhook-emission",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P v2.2 \u2014 webhook ENTRY/EXIT emission into alerts_tradingview.indicator_alerts",
      "property": "alerts-hotkeys-etc",
      "project": "pinepb",
      "sprint": 3,
      "created_at": "2026-06-26",
      "path": "trading/indicators/src/indicators/J3P_v2.2_SIP_Webhook.pine",
      "note": "SHIPPED 2026-06-26: FashLate-shape alert() JSON for all 3 plays (VWAP Pull-In/Trend/Backside) + per-play ENTRY->stop/target/EOD lifecycle. Live smoke test tracked in j3p-webhook-smoke-test.",
      "completed_at": "2026-06-26"
    },
    {
      "id": "j3p-webhook-smoke-test",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P \u2014 live smoke test: TV alert -> webhook -> indicator_alerts row (1 SIP name)",
      "property": "alerts-hotkeys-etc",
      "project": "pinepb",
      "sprint": 3,
      "created_at": "2026-06-26",
      "note": "Needs James: attach 'Any alert() function call' alert with webhook -> tradingview_smb_indicator_trigger.php?key=<shared_secret>. Confirm indicator='J3P' row + entry/exit prices."
    },
    {
      "id": "j3p-label-alert-polish",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "J3P \u2014 FashLate-style label/alert polish (compact toggle + per-play templates)",
      "property": "playbooks",
      "project": "pinepb",
      "sprint": 3,
      "created_at": "2026-06-26"
    },
    {
      "id": "j3p-trend-play-rebuild-v23",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "J3P v2.3 \u2014 rebuild Trend play as Dale VWAP 1st-deviation pullback (retire 9-EMA trigger)",
      "property": "playbooks",
      "project": "pinepb",
      "sprint": 3,
      "created_at": "2026-06-26",
      "note": "Per TraderDale_VWAP.pine + VWAP Book: Dale = 1st-dev only, stop = Reaction Point."
    },
    {
      "id": "sip-deploy-j3p-watchlist",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Deploy J3P across SIP master_watchlist (TV layout + webhook alert per symbol)",
      "property": "alerts-hotkeys-etc",
      "project": "sip",
      "sprint": 3,
      "created_at": "2026-06-26",
      "note": "2026-06-26 master_watchlist: CALY SOXL LLY CASY VMI BH AYI BE PRAX FTAI SQQQ."
    },
    {
      "id": "sip-watchlist-tv-autosync",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "title": "Auto-sync /sip/watchlist.json -> TradingView symbol list for J3P coverage rotation",
      "property": "alerts-hotkeys-etc",
      "project": "sip",
      "sprint": 3,
      "created_at": "2026-06-26"
    },
    {
      "id": "exec-backtest-j3p-plays",
      "phase": "phase-2_5",
      "status": "in_progress",
      "owner": "claude",
      "priority": "P1",
      "title": "Backtest J3P plays (Focus-3/KITE) on SIP names \u2014 edge by playbook x grade x tier before hotkeys",
      "property": "param_surface",
      "project": "exec",
      "sprint": 4,
      "created_at": "2026-06-26",
      "note": "v1 6/27 (11 SIP names, n=157). 6/28 EXPANDED to WIDE 52-name canonical (n=804) + James daily universe (37 names, n=411) + v2.3 (Dale-faithful + CBO expansion) variants -> 11 registered runs (/system/J3P_RUNS.html). EDGE FLIP: yesterday's 'Trend = only +edge' did NOT survive out-of-sample -- on WIDE, Trend = -0.282R, every play loses -0.16..-0.38R, and v2.3/CBO made it WORSE everywhere (WIDE -15180 -> -18420). Entry-model ruled out ('it's not the entry'). New hypothesis: regime filtering (trade only when tape is trending, in that direction) + ADD SHORTS (system is long-only) + deeper data (kti caps ~14d). v2.4 filter matrix scaffolded (j3p_v24_matrix.py: shorts/rot-block/market-regime/Shannon-40MA/grade-A-B/combo) but runs PENDING. Plan published /system/J3P_EXPERIMENT_PLAN.html. Engine Z:\\trading\\dev\\j3p-backtest\\. Next: run v2.4 matrix + stand up deep-history adapter (Polygon/Databento)."
    },
    {
      "id": "exec-j3p-backtest-v1-viewers-2026-06-27",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P backtest v1 engine + edge read + interactive chart viewers published",
      "property": "param_surface",
      "project": "exec",
      "sprint": 4,
      "created_at": "2026-06-27",
      "path": "/system/J3P_BACKTEST_EDGE.html",
      "note": "Saturday build. Engine Z:\\trading\\dev\\j3p-backtest\\ (j3p_backtest.py + make_report.py + j3p_charts.py; real karb 1-min->5-min bars, $100 constant risk/trade, VCET-conditioned). 11 SIP names x 10d = 157 trades. Published 6 surfaces to /system/: J3P_BACKTEST_EDGE (verdict), _MARKET, _CHARTS, _MARKET_CHARTS, _VIEWER (lightweight-charts interactive), _WIDE. Verdict: Trend(Dale 1st-dev) only +edge (+$361/+0.078R, grade B +0.383R/75%); Pull-In -$2410 with 94/106 fills in ROTATION (regime-gate it = VCET expansion thesis confirmed live); Backside n=4 unproven. Stage-4 edge gate: Trend/B-C/TREND_UP advances; Pull-In only once regime-gated; Backside does not.",
      "completed_at": "2026-06-27"
    },
    {
      "id": "exec-j3p-wide-sample-registry-experiment-plan-2026-06-28",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P backtest expanded to WIDE 52-name canonical + runs registry + v2.4 experiment plan",
      "property": "param_surface",
      "project": "exec",
      "sprint": 4,
      "created_at": "2026-06-28",
      "path": "/system/J3P_RUNS.html",
      "note": "Sunday research synthesis. Expanded the biased 11-name SIP cut into a credible sample: WIDE 52-name canonical (n=804, 12d), James daily universe (37 names traded 6/12-6/24, n=411), Market-18, and v2.3 (Dale-faithful pullback + CBO expansion) variants of each = 11 registered runs. New tooling: j3p_runs_registry.py (-> J3P_RUNS.html/json head-to-head), j3p_diagnose.py, j3p_export_viewer.py (-> J3P_SETUPS/J3P_BACKTEST_VIEWER/j3p-viewer), j3p_sip_vs_bucket.py (-> J3P_SIP_VS_BUCKET.html), j3p_v24_matrix.py (v2.4 filter-matrix scaffold). HEADLINE FINDING: the Trend +edge from v1 did NOT survive out-of-sample -- WIDE Trend = -0.282R (was +0.078R on SIP-11), every play loses -0.16..-0.38R, v2.3/CBO made it worse everywhere (WIDE -15180 -> -18420). Entry-model ruled out. Confirms the recurring thesis (Focus-3): selectivity is the only positive lever (A-grade +0.17R vs C -0.34R). Experiment plan published /system/J3P_EXPERIMENT_PLAN.html: (A) add shorts, (B) VCET rotation-block, (C) market-regime gate, (D) Shannon 40-MA, grade A/B gate, skip-midday, + deep-history data adapter (kti caps ~14d).",
      "completed_at": "2026-06-28"
    },
    {
      "id": "exec-hotkeys-j3p-setups",
      "phase": "phase-2_5",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Wire setups.today.json -> hotkey-engine for J3P-graded entries (SIM -> forward-test)",
      "property": "alerts-hotkeys-etc",
      "project": "exec",
      "sprint": 4,
      "created_at": "2026-06-26",
      "note": "hotkey-engine V1 SIM done; grade A/B/C/D = 30/20/10/5% daily stop; ConfirmationGate + post-fill flatten watchdog; no fire-on-keystroke."
    },
    {
      "id": "exec-kore-live-j3p",
      "phase": "phase-2_5",
      "status": "blocked",
      "owner": "claude",
      "priority": "P2",
      "title": "Automate J3P execution via Kore backend (firm-confirm Gr8Script -> forward -> live)",
      "property": "alerts-hotkeys-etc",
      "project": "exec",
      "sprint": 5,
      "created_at": "2026-06-26",
      "note": "Blocked: firm-confirm Gr8Script/Kore call signatures to un-inert kore_backend.py. Server-side managed stop so fills are never naked."
    },
    {
      "id": "sip-g8trade-roadmap-plan",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "PLAN published: SIP->Indicators->g8trade roadmap",
      "property": "pm_system",
      "project": "exec",
      "sprint": 3,
      "created_at": "2026-06-26",
      "path": "/system/SIP_INDICATORS_G8TRADE_PLAN.md",
      "note": "Published 2026-06-26 to /system/ hub. Stages 1-5 with edge gate before hotkeys.",
      "completed_at": "2026-06-26"
    },
    {
      "id": "sip-mgp-explained-page-2026-06-23",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "SIP/MGP explainer page published",
      "property": "mgp",
      "project": "sip",
      "created_at": "2026-06-23",
      "completed_at": "2026-06-23",
      "path": "/system/sip-mgp-explained.html",
      "note": "Published 2026-06-23: /system/sip-mgp-explained.html explainer documenting the SIP->MGP pipeline. _shared/jd_nav.js + system/all-pages.html sitemap regenerated to link it."
    },
    {
      "id": "exec-j3p-v24-filter-matrix-run-2026-06-29",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P v2.4 regime-filter matrix run end-to-end (shorts + Dale-trend + CBO + 6 filters)",
      "property": "param_surface",
      "project": "exec",
      "sprint": 4,
      "created_at": "2026-06-29",
      "path": "/system/J3P_V24_MATRIX.html",
      "note": "Yesterday's #1 focus, executed. WIDE 52-name ~14d window. Verdict: filtering narrows the bleed but does NOT create edge -- best variant grade A/B + Shannon-40 at -0.120R vs v2.3 baseline -0.194R. Shorts added (CBO-short + Dale-trend-short): short side -0.144R, less-bad than longs but still negative. COMBINED 'everything lines up' = 127 trades / -0.193R. Conclusion: entry-model question stays closed; the gate is now DEEPER DATA -- kti's ~14d cap can't judge a trend system. The minute-bar adapter (exec-backtest-j3p-plays) is the real blocker.",
      "completed_at": "2026-06-29"
    },
    {
      "id": "nightly-j3p-parity-2min-2026-07-01",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-01",
      "owner": "James",
      "priority": "P2",
      "property": "system",
      "title": "J3P 2-min parity build + TOD matrix page",
      "path": "/system/J3P_PARITY_2MIN.html",
      "note": "Built j3p-parity-2min tree + J3P_PARITY_2MIN.html + J3P_TOD_MATRIX.html."
    },
    {
      "id": "nightly-j3p-exit-surface-2026-07-02",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-02",
      "owner": "James",
      "priority": "P2",
      "property": "system",
      "title": "J3P exit surface + ladder OOS + Data Trust Spec refresh",
      "path": "/system/J3P_EXIT_SURFACE.html",
      "note": "J3P_EXIT_SURFACE.html, J3P_LADDER_OOS.html, DATA_TRUST_SPEC.html updated."
    },
    {
      "id": "nightly-vwap-reject-short-playbook-2026-07-04",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-04",
      "owner": "James",
      "priority": "P2",
      "property": "playbooks",
      "title": "vwap_reject_short playbook page",
      "path": "/system/J3P_PLAYBOOK_VWAP_REJECT_SHORT.html",
      "note": "Sat build: playbook page for the one durable OOS play (vwap_reject_short)."
    },
    {
      "id": "nightly-sip-nuance-tracking-2026-07-06",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-06",
      "owner": "James",
      "priority": "P1",
      "property": "system",
      "title": "SIP Today + Nuance tracking + Exit study + Trades-by-day + Replay tool",
      "path": "/system/J3P_SIP_TODAY.html",
      "note": "J3P_SIP_TODAY, J3P_NUANCE, J3P_EXIT_STUDY, J3P_TRADES_BY_DAY(+csv), j3p-replay tree, 3 playbook pages, hub+playbooks index."
    },
    {
      "id": "nightly-j3p-visual-stack-2026-07-07",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-07",
      "owner": "James",
      "priority": "P1",
      "property": "system",
      "title": "J3P visual stack: cheat cards, swing DMA, trend playbook, analog target, trust report, alerts settings",
      "path": "/system/J3P.html",
      "note": "~25 pages: J3P_CHEAT, J3P_SWING_DMA, J3P_TREND_PLAYBOOK, J3P_ANALOG_TARGET, J3P_TRUST_REPORT, J3P_BRACKET, J3P_TRADE_CHECK, 5x J3P_REPLAY_*, alerts-settings/config+banner.js, J3P hub."
    },
    {
      "id": "nightly-j3p-watcher-2026-07-08",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-08",
      "owner": "James",
      "priority": "P2",
      "property": "system",
      "title": "J3P watcher page + alerts cheat card",
      "path": "/system/J3P_WATCHER.html",
      "note": "J3P_WATCHER.html + J3P_ALERTS_CHEAT.html. NOTE: app-status/freshness monitor last ran 12:01 this day and then stopped."
    },
    {
      "id": "nightly-j3p-live-setups-2026-07-09",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-09",
      "owner": "James",
      "priority": "P1",
      "property": "scorecards",
      "title": "J3P live setups + churn-today meter + command surface + multi-tf-stack",
      "path": "/system/J3P_LIVE.html",
      "note": "live_setups.json + churn_today.json auto-populated, J3P_LIVE.html, J3P_EXPERIMENT_LOG.html, command.html refresh, multi-tf-stack tree, atlas-flow-status/full."
    },
    {
      "id": "nightly-vwap-band-research-system-2026-07-10",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "VWAP Band Research System dashboard shipped",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "property": "monthly_reports",
      "completed_at": "2026-07-10",
      "created_at": "2026-07-10",
      "note": "New research surface at smb.a1prospect.com/db_reports/vwap-research/. Core spec: entry VWAP -> -0.5SD, stop LB1, target UB3 (+2.0SD). Compression/expansion analysis (BB squeeze 20/2SD, ATR-14 52wk percentile, VWAP SD percentile, BB %B, HV 5d/20d) with confirmed/partial/mixed/not_confirmed verdict boxes. Ties to VWAP Band Entry Surface + Param Surface research."
    },
    {
      "id": "nightly-atlas-flow-status-2026-07-10",
      "phase": "phase-0",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Atlas flow-status pipeline verifier (a2) shipped",
      "path": "/z/public_html/system/atlas-flow-status.html",
      "property": "foundation",
      "completed_at": "2026-07-10",
      "created_at": "2026-07-10",
      "note": "New end-to-end flow verifier for the a2 scan/SIP/MGP pipeline: 16 checks across preflight/scanners/sip/multitf/mgp stages -> atlas-flow-status.{html,json} + atlas-flow-full.{html,json}. First run 13 pass / 2 fail / 1 skip = RED (PBS pullback scanner FAIL; V1 recap flagged but recap MGP actually emailed 16:30 per notify.log -> likely snapshot-timing). NOTE: this covers the a2 pipeline; it does NOT replace the dead host-level app-status/watchdog (still frozen 7/8 12:01)."
    },
    {
      "id": "nightly-sip-briefing-pipeline-refresh-2026-07-10",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "SIP/briefing generator pipeline refresh (7 modules)",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "property": "mgp",
      "completed_at": "2026-07-10",
      "created_at": "2026-07-10",
      "note": "Touched fetch_briefing_catalysts.py, build_sip_page.py, hydrate_companies.py (+companies_cache), market_sentiment.py, session_verdict.py, hydrate_liquidity.py (+liquidity_cache), sip_canonical_writer.py, run_sip_build_a2.py between 16:05-17:15. SIP nuance + briefing catalyst hydration path."
    },
    {
      "id": "earn-gap-reclaim-massive-validated-2026-07-11",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Earnings gap-down + VWAP-reclaim play validated on FULL Massive universe \u2014 first edge to pass every honesty check",
      "property": "param_surface",
      "project": "exec",
      "path": "/system/J3P_EXPERIMENT_LOG.html",
      "completed_at": "2026-07-11",
      "created_at": "2026-07-11",
      "note": "James's own play (gap down \u2265-2% vs close \u2192 break AND HOLD session VWAP \u2192 target 2-day AVWAP, stop under LOD). Built gap_reclaim.py, tested in 3 iterations same day: (1) wrong spec (prior-day-VWAP magnet) DIDN'T hold (26% touch); (2) corrected v2 spec PARTIAL (43% touch, ~breakeven, only 35 earnings setups testable in SIP-only bar cache \u2014 which IS James's point that SIP misses earnings names); (3) FULL universe via Massive UNLIMITED upgrade = 686 earnings gap-downs \u2265-2% + liquid over 87 days. Result: WORKS and NOT tail-driven \u2014 513/686 fired reclaim+hold (75%); avgR +0.124 with POSITIVE median (+0.068), 53% win, t=2.55; BOTH halves positive (train +0.09 / test +0.26); positive 4 of 5 months incl. falling June (+0.18). Gap-size is the dial: >5% = +0.29R avg/57% win/t=3.8 (grade A); 3\u20135% flat (B); 2\u20133% negative (C, skip). First J3P play to clear look-ahead + both-halves + not-tail-driven at once. The Massive unlimited fetch is what broke the ~2-week/~14d data wall (see reference_massive_api / deep_data_massive)."
    },
    {
      "id": "j3p-live-earnings-scan-2026-07-11",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Live earnings scan wired into J3P scanner + SIP page (earn_gap_reclaim LAB play)",
      "property": "sip",
      "project": "sip",
      "path": "/system/J3P_SIP_TODAY.html",
      "completed_at": "2026-07-11",
      "created_at": "2026-07-11",
      "note": "Liquid earnings-morning names (kti news earnings: prior-day after-hours + same-day pre-market) now auto-append to the scanner watchlist every morning \u2014 directly addresses 'SIP misses earnings names'. earn_gap_reclaim deployed to the live scanner as a LAB play (morning-gated, trend-gate exempt) so the forward tracker measures James's real version. On the SIP page: earnings badge + per-name modal + dedicated earnings section. Every fire alerts Discord/SMS/banner with the gap grade (A/B/C by gap size). J3P_LIVE.html + J3P_SIP_TODAY.html + J3P_EXPERIMENT_LOG.html regenerated."
    },
    {
      "id": "scorecard-chart-click-precompute-2026-07-11",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "Scorecard chart-click (SetupChart modal) rebuilt + full precompute regen of all trade scorecards + You-vs-You refresh",
      "property": "scorecards",
      "project": "acct",
      "path": "/scorecards/",
      "completed_at": "2026-07-11",
      "created_at": "2026-07-11",
      "note": "Click a symbol on a scorecard \u2192 SetupChart modal with entry/exit pennants. Reworked scorecard_charts.js (both /_shared/ and a /scorecards/ local copy \u2014 the /_shared/ 404-on-smb gotcha) + setup_chart.js + manifest.json, then bulk-regenerated every trade_scorecard_*.html (Apr 10 \u2192 Jul 11) so the click-through is live across the whole archive. You-vs-You self-benchmark scorecard also refreshed (last 15 trading days, 6/12\u21927/10)."
    },
    {
      "id": "lynch-combinatorial-backtest-2026-07-12",
      "phase": "design",
      "status": "completed",
      "completed_at": "2026-07-12",
      "title": "LYNCH Continuation Breakout combinatorial backtest suite shipped",
      "owner": "James",
      "priority": "P1",
      "property": "backtests",
      "note": "12-page study: LYNCH_MASTER (master combinatorial), BACKTEST (data-driven rule rec), 5YR, WIDE, 5YR_SECTOR_RRG, TRADES, DOLLARS, DD (drawdown), SENTIMENT, DAYTRADE, HYBRID, ARCHIVE. Under /system/LYNCH_*.html.",
      "path": "/system/LYNCH_MASTER.html"
    },
    {
      "id": "chart-selfhost-vendor-2026-07-12",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-12",
      "title": "Self-hosted charting vendor lib into /_shared/vendor + rewired setup_chart.js",
      "owner": "James",
      "priority": "P1",
      "property": "_shared",
      "note": "Added /_shared/vendor, kept setup_chart.js.pre-selfhost.bak, rewrote setup_chart.js to load the vendor bundle locally. Targets the /_shared 404s that break charts cross-origin on smb.",
      "path": "/_shared/setup_chart.js"
    },
    {
      "id": "j3p-earnings-replay-2026-07-12",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-12",
      "title": "J3P Earnings Pipeline Replay viewer + J3P page regen",
      "owner": "James",
      "priority": "P1",
      "property": "system",
      "note": "New /system/J3P_EARNINGS_REPLAY.html for the earn_gap_reclaim play; regenerated J3P.html, J3P_LIVE, J3P_LIVE_DRILL, J3P_SIP_TODAY, J3P_SIP_DEMO, J3P_EXPERIMENT_LOG, J3P_RUNBOOK.",
      "path": "/system/J3P_EARNINGS_REPLAY.html"
    },
    {
      "id": "scorecards-regen-chartclick-2026-07-12",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-12",
      "title": "Scorecards regenerated with self-hosted chart-click (7/12)",
      "owner": "James",
      "priority": "P2",
      "property": "scorecards",
      "note": "trade_scorecard_2026-07-12 + manifest + you-vs-you refreshed against the new /_shared/vendor chart bundle; J3P_EXECUTION_REPORT regenerated.",
      "path": "/scorecards/you-vs-you.html"
    },
    {
      "id": "sip-builder-earnings-integration-2026-07-12",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-12",
      "title": "SIP page builder + picks enrichment updated for earnings integration",
      "owner": "James",
      "priority": "P1",
      "property": "system",
      "note": "build_sip_page.py reworked to fold earnings gappers into the SIP page (kept build_sip_page.py.pre-earnings.bak; _sip_preview_v2 regenerated); enrich_picks.py updated. Part of the earn_gap_reclaim / earnings-into-SIP pipeline.",
      "path": "/system/J3P_SIP_TODAY.html"
    },
    {
      "id": "lynch-earnings-pit-live-surface-2026-07-13",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "LYNCH earnings + PIT live surface (TRUERULES) built on Sunday's combinatorial backtest suite",
      "path": "/system/LYNCH_PIT_TRUERULES.html",
      "completed_at": "2026-07-13",
      "created_at": "2026-07-13",
      "note": "Turned Sunday's LYNCH Continuation-Breakout backtest suite into a live decision surface: LYNCH_MASTER (07:46), LYNCH_EARNINGS (12:23), LYNCH_ARCHIVE (13:00), LYNCH_PIT (18:58) + LYNCH_PIT_TRUERULES (19:18). PIT = the live pit read; TRUERULES = the codified rule set the backtest validated."
    },
    {
      "id": "setup-chart-price-markers-2026-07-13",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "title": "setup_chart.js: entry/exit price markers on annotated charts",
      "path": "/_shared/setup_chart.js",
      "completed_at": "2026-07-13",
      "created_at": "2026-07-13",
      "note": "Added price markers to the shared annotated-chart renderer (kept setup_chart.js.pre-pricemarkers.bak). Feeds every scorecard + LYNCH/J3P surface that uses setup_chart.js. Regenerated scorecards, you-vs-you, J3P_LIVE against it same day."
    },
    {
      "id": "market-sentiment-visual-rework-2026-07-13",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "market_sentiment.py alignment + visual rework",
      "path": "/z/setup_daily_reporting/reports/market_sentiment.py",
      "completed_at": "2026-07-13",
      "created_at": "2026-07-13",
      "note": "Two-pass rework of the sentiment generator: .bak-align (16:08) then .bak-visual then final (19:25). Drives the MGP sentiment block + STRATEGY_SCOREBOARD. Also touched alpha_trend.py (08:49), build_sip_page.py (09:02), enrich_picks.py (12:33), freshness_board.py (15:17) same day."
    },
    {
      "id": "j3p-monday-live-test-surfaces-2026-07-13",
      "phase": "phase-2_5",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P Monday live-test surfaces regenerated (LIVE, WATCHER, ROBOT_REPORT, churn/live_setups)",
      "path": "/system/J3P_LIVE.html",
      "completed_at": "2026-07-13",
      "created_at": "2026-07-13",
      "note": "First live session since the earn_gap_reclaim scan went in Sunday. Regenerated J3P_SIP_DEMO (07:43), J3P_EARNINGS_REPLAY (07:50), J3P_EXPERIMENT_LOG (07:52), J3P_WATCHER (08:01), J3P_SIP_TODAY (09:10), J3P_LIVE + J3P_LIVE_DATA (16:15), J3P_ROBOT_REPORT (16:58); scorecards churn_today + live_setups (16:04). Pipeline ran clean; whether the live GO actually fired on a held VWAP reclaim still needs eyeball confirmation from J3P_LIVE_DATA / the tracker."
    },
    {
      "id": "atlas-flow-status-monitor-live-2026-07-13",
      "phase": "phase-project-mgmt",
      "status": "completed",
      "owner": "claude",
      "title": "Atlas flow-status monitor live and fresh (candidate replacement for dead app-status watchdog)",
      "path": "/system/atlas-flow-status.html",
      "completed_at": "2026-07-13",
      "created_at": "2026-07-13",
      "note": "atlas-flow-status.{html,json} + atlas-flow-full.{html,json} regenerated 20:40 today (fresh), alongside a new freshness_board.py (15:17) + freshness.html (17:55). Meanwhile the OLD app-status.json is STILL frozen at 2026-07-08 12:01:21 (6th day). Open question for 7/14: is atlas-flow the intended replacement so app-status can be retired, or must the old watchdog be revived?"
    },
    {
      "id": "jd-earnings-pattern-analysis-2026-07-14",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Earnings-trade pattern analysis \u2014 reverse-engineer the filters the backtest was missing",
      "path": "/z/public_html/system/JD_EARNINGS_PATTERN.html",
      "completed_at": "2026-07-14",
      "created_at": "2026-07-14",
      "property": "scorecards",
      "note": "Built JD_EARNINGS_PATTERN.html from the 180-day fill log (18,456 fills; 1,191 earnings-adjacent; -$631; 47.0% win vs 44.1% overall). Winners-vs-losers median split + by-hour/by-side/days-to-earnings breakdowns. Actionable output: skip-list candidates (SMCI -$137, NVDA -$116, AMD -$72, CRWV -$82), earnings winners (TSLA/AMZN/MU), best hour 17:00 (+$116), day-of highest win rate (52%). Feeds exec-backtest-j3p-plays selection filters."
    },
    {
      "id": "j3p-tuesday-live-surfaces-2026-07-14",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "J3P Tuesday live-test surface set regenerated (2nd live session)",
      "path": "/z/public_html/system/J3P_LIVE.html",
      "completed_at": "2026-07-14",
      "created_at": "2026-07-14",
      "property": "live_classifier",
      "note": "Second live session since the earn_gap_reclaim scan went in. Regenerated J3P_LIVE + J3P_LIVE_DATA + J3P_WATCHER (08:01) + J3P_ROBOT_REPORT (16:15) + J3P_SIP_TODAY (20:19); scorecards churn_today + live_setups (16:04), trade_scorecard_2026-07-14 + you-vs-you (16:47-48); multi-tf-stack refresh (20:31). GO-firing still needs the same eyeball confirmation carried from Monday."
    },
    {
      "id": "j3p-wednesday-live-surfaces-2026-07-15",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "J3P Wednesday live-test surface set regenerated (3rd live session) + LYNCH archive rebuild",
      "path": "/system/J3P_LIVE.html",
      "completed_at": "2026-07-15",
      "created_at": "2026-07-15",
      "property": "live_classifier",
      "note": "Third consecutive live session for the earn_gap_reclaim scan. Regenerated J3P_WATCHER (08:01), J3P_LIVE + J3P_LIVE_DATA (16:16), J3P_ROBOT_REPORT (16:17), J3P_LIVE_DRILL (19:40), J3P_SIP_TODAY (20:12), command.html (20:24), multi-tf-stack (20:31); scorecards churn_today + live_setups (16:04), trade_scorecard_2026-07-15 + manifest + you-vs-you (16:46), lean_preview 7/14 + 7/15 (17:30). LYNCH_ARCHIVE rebuilt (13:00) \u2014 forward-test record now 3 days / 4 picks / 3 SIP hits / 0 outcomes computed yet. All 3 MGP editions emailed clean (morning 05:42+06:49 risk-on/breadth+1, midday 11:30+11:38 yellow-lights, recap 16:31+16:41 chop day). atlas-flow-FULL monitor GREEN today (20:30, 13/13). No net-new source commits in reports/, accountability-v2, scripts, indicators, ddl or db_reports trees \u2014 output/pipeline day. Live GO eyeball confirmation still owed across all 3 sessions."
    },
    {
      "id": "sip-page-generator-update-2026-07-16",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "SIP page generator (build_sip_page.py) updated",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "completed_at": "2026-07-16",
      "created_at": "2026-07-16",
      "note": "build_sip_page.py edited 16:53. Server-side scan confirmed it as the only reports/-tree source change across 07-16..07-18 (otherwise an output/pipeline period). Alongside published J3P_COMPRESSION_REPORT (01:15) + J3P_LIVE_DRILL (16:51).",
      "property": "mgp",
      "project": "atlas"
    },
    {
      "id": "hud-prod-trader-php-edit-2026-07-16",
      "phase": "phase-4",
      "status": "completed",
      "owner": "james",
      "title": "Prod HUD api/trader.php edited (backup tar taken)",
      "path": "/z/trading/a1p_journal/accountability-v2/api/trader.php",
      "completed_at": "2026-07-16",
      "created_at": "2026-07-16",
      "note": "Prod accountability-v2/api/trader.php touched 21:06; a2 has backup_hud_trader_php_20260716.tar from the same day, so a deliberate prod HUD edit was made with a backup. REPORTED not edited (HUD is off-limits to Claude). Worth confirming intent \u2014 trader.php is the endpoint behind the HUD LIMIT-100 closed-trades truncation note.",
      "property": "accountability_hud",
      "project": "acct-v2"
    },
    {
      "id": "vwap-research-page-2026-07-17",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "james",
      "title": "VWAP research page (db_reports/vwap-research) updated",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "completed_at": "2026-07-17",
      "created_at": "2026-07-17",
      "note": "db_reports/vwap-research/index.html rebuilt 16:46 \u2014 the only db_reports/ddl change across the 3-day window. Same day the full J3P live pipeline ran clean (J3P_WATCHER 08:00, J3P_SIP_TODAY 08:45, J3P_LIVE 16:16, ROBOT_REPORT 16:16, command 16:50, LYNCH_ARCHIVE 17:00, multi-tf-stack 20:31) and atlas-flow-FULL ran GREEN (20:39).",
      "property": "strategy_library",
      "project": "stratlib"
    },
    {
      "id": "hud-dev-multitrader-session-2026-07-18",
      "phase": "phase-4",
      "status": "completed",
      "owner": "james",
      "title": "Dev-HUD multi-trader work session (traders.php, trader.php, render.js)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/",
      "completed_at": "2026-07-18",
      "created_at": "2026-07-18",
      "note": "Saturday (market closed). Real activity was dev-HUD editing: accountability-v2/api/traders.php (11:38), api/trader.php (17:06), hud/js/render.js (17:49) in the DEV tree \u2014 multi-trader endpoint + HUD render work. REPORTED not edited (HUD off-limits). Ties to the in-progress scorecards-multitrader-migration. Otherwise a quiet weekend: scorecards + all 3 MGP editions only (MGP fired LATE ~20:55 \u2014 machine-slept batch).",
      "property": "accountability_hud",
      "project": "acct-v2"
    },
    {
      "id": "lynch-master-combinatorial-backtest",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "LYNCH Master combinatorial backtest lab (3,672 combos, walk-forward validated)",
      "path": "/z/public_html/system/LYNCH_MASTER.html",
      "property": "backtests",
      "completed_at": "2026-07-19",
      "created_at": "2026-07-19",
      "note": "New combinatorial backtest surface built Sun 22:00. Grades \u00d7 Sizing \u00d7 Exit \u00d7 Regime-gate = 3,672 combos, each a 5-yr real-dollar sim (whole shares, $50k position cap, SMB commissions). 1147/3672 survived both-halves walk-forward. Best return-per-dollar-of-pain (and best raw $): A+B+C grade \u00b7 UVXY-mom\u00d7Reactive sizing \u00b7 Swing pure (hold 5d) exit \u00b7 no gate \u2014 435 trades, +$71,454, $14,992 max DD, 477% return/DD, 45.1% win, Sharpe 0.74. Includes a plain-English '10-year-old report'. Sibling pages: LYNCH_HYBRID / LYNCH_SENTIMENT / LYNCH_DOLLARS / LYNCH_DD."
    },
    {
      "id": "j3p-week-ahead-2026-07-20",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "J3P Week Ahead \u2014 Monday 2026-07-20 prep surface",
      "path": "/z/public_html/system/J3P_WEEK_AHEAD.html",
      "property": "system",
      "completed_at": "2026-07-19",
      "created_at": "2026-07-19",
      "note": "Sunday auto-rebuild (11:51). Forward-looking prep for Mon 7/20: Friday recap table (earn_gap_reclaim +5.2R best, vwap_pullin/revbreak the drags) + 21 names to prepare, ordered Monday reporters (DPZ earnings pre-market) \u2192 SIP \u2192 Ants \u2192 Friday movers. Each card: 5-DMA lean side, catalyst/scanner source, VWAP-pullin play, Fri close vs 5-DMA, click-to-chart with TF + MA toggles."
    },
    {
      "id": "cross-asset-multitf-engine-edits-2026-07-19",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Cross-asset / multi-TF-stack engine edits (xa + gate + all-day launcher)",
      "path": "/z/setup_daily_reporting/reports/gen_multi_tf_stack.py",
      "property": "system",
      "completed_at": "2026-07-19",
      "created_at": "2026-07-19",
      "note": "Sun 10:10-11:01 source work on the overnight cross-asset / multi-TF-stack pipeline: cross_asset_panel.py (10:10), gen_multi_tf_stack.py edited twice (10:12 .bak-xa cross-asset, 10:48 .bak-gate), start_multi_tf.cmd (11:01 .bak-allday launcher). Feeds the system/multi-tf-stack surface that refreshed 20:00."
    },
    {
      "id": "build-sip-page-nav-fix-2026-07-19",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "SIP page builder nav fix (build_sip_page.py)",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "property": "system",
      "completed_at": "2026-07-19",
      "created_at": "2026-07-19",
      "note": "Sun 11:51 edit (.bak-nav-0719) to the SIP page builder \u2014 nav fix; coincides with the J3P_WEEK_AHEAD 11:51 rebuild and _sip_preview_v2 refresh."
    },
    {
      "id": "market-sentiment-leans-edit-2026-07-19",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "title": "Market-sentiment engine 'leans' edit (market_sentiment.py)",
      "path": "/z/setup_daily_reporting/reports/market_sentiment.py",
      "property": "mgp",
      "completed_at": "2026-07-19",
      "created_at": "2026-07-19",
      "note": "Sun 19:42 edit (.bak-leans) to market_sentiment.py \u2014 the sentiment/leans engine feeding MGP posture and the Week-Ahead lean chips."
    },
    {
      "id": "mgp-v2-generator-refresh-0720",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "property": "mgp",
      "title": "MGP v2 generator + index refresh",
      "path": "/z/setup_daily_reporting/reports/gen_mgp_v2.py",
      "completed_at": "2026-07-20",
      "created_at": "2026-07-20",
      "note": "gen_mgp_v2_index.py (08:42) + gen_mgp_v2.py (08:44) both edited. All three MGP editions published and emailed the same day."
    },
    {
      "id": "sip-page-stale-guard-0720",
      "phase": "phase-sip",
      "status": "completed",
      "owner": "claude",
      "property": "mgp",
      "title": "SIP page: stale-guard + strip-details pass",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "completed_at": "2026-07-20",
      "created_at": "2026-07-20",
      "note": "Two checkpointed edits to build_sip_page.py: .bak-staleguard-0720 (09:16) then .bak-stripdetails-0720 (09:35). Stale-data guard so J3P_SIP_TODAY cannot silently show yesterday's names."
    },
    {
      "id": "sip-page-jsfix-mover-0721",
      "phase": "phase-sip",
      "status": "completed",
      "owner": "claude",
      "property": "mgp",
      "title": "SIP page: JS fix + mover-detection pass",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "completed_at": "2026-07-21",
      "created_at": "2026-07-21",
      "note": ".bak-jsfix-0721 (09:04) then build_sip_page.py + .bak-mover-0721 (12:43). Second day of consecutive SIP-page hardening."
    },
    {
      "id": "j3p-live-drill-surface",
      "phase": "phase-j3p",
      "status": "completed",
      "owner": "claude",
      "property": "alerts-hotkeys-etc",
      "title": "J3P Live Drill surface published (page + data feed)",
      "path": "/z/public_html/system/J3P_LIVE_DRILL.html",
      "completed_at": "2026-07-21",
      "created_at": "2026-07-21",
      "note": "J3P_LIVE_DRILL_DATA.json + J3P_LIVE_DRILL.html written together at 12:42 \u2014 practice/drill surface alongside the live J3P stack."
    },
    {
      "id": "james-daily-review-whatif-0721",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "monthly_reports",
      "title": "What-if daily review page for 2026-07-21",
      "path": "/z/gt_dev/db_reports/james-daily-reviews/james_2026-07-21_whatif.html",
      "completed_at": "2026-07-21",
      "created_at": "2026-07-21",
      "note": "New james-daily-reviews/*_whatif.html artifact (19:39) \u2014 counterfactual review of the day's fills."
    },
    {
      "id": "mgp-tone-spec-published",
      "phase": "phase-mgp",
      "status": "completed",
      "owner": "claude",
      "property": "mgp",
      "title": "MGP tone spec published (locked trader-brief voice)",
      "path": "/z/public_html/system/MGP_TONE_SPEC.md",
      "completed_at": "2026-07-22",
      "created_at": "2026-07-22",
      "note": "MGP_TONE_SPEC.md written 20:14 \u2014 the locked voice for all three MGP editions (verdict -> snapshot -> catalysts -> read -> watch)."
    },
    {
      "id": "system-all-pages-rebuild-0722",
      "phase": "phase-foundation",
      "status": "completed",
      "owner": "claude",
      "property": "foundation",
      "title": "System hub all-pages index rebuilt",
      "path": "/z/public_html/system/all-pages.html",
      "completed_at": "2026-07-22",
      "created_at": "2026-07-22",
      "note": "all-pages.html regenerated 07:44 with a .bak-2026-07-21 checkpoint taken the night before."
    },
    {
      "id": "exec-scorecard-v3-inject-start",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "scorecards",
      "title": "Execution Scorecard v3: first injection into a live scorecard",
      "path": "/z/public_html/scorecards/scorecard_v3_2026-07-24.html",
      "completed_at": "2026-07-24",
      "created_at": "2026-07-24",
      "note": "chart_assets/ + trade_scorecard_2026-07-24.html.pre_v3_inject.bak at 23:46 \u2014 the v3 campaign-first rebuild starts landing in the published scorecards."
    },
    {
      "id": "scorecard-effective-stop-enrichment",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "scorecards",
      "title": "Scorecard enrichment: effective-stop derivation",
      "path": "/z/setup_daily_reporting/reports/scorecard_enrichments.py",
      "completed_at": "2026-07-25",
      "created_at": "2026-07-25",
      "note": "scorecard_enrichments.py + .bak_pre_effective_stop at 08:27 \u2014 adds an effective-stop field so v3 can grade risk against the stop actually in force."
    },
    {
      "id": "v3-eod-runner",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "scorecards",
      "title": "v3 EOD runner (v3_eod.sh) split out of run_all_eod",
      "path": "/z/setup_daily_reporting/reports/v3_eod.sh",
      "completed_at": "2026-07-25",
      "created_at": "2026-07-25",
      "note": "run_all_eod.bat.bak_pre_v3 (18:46) then v3_eod.sh (19:51). Dedicated end-of-day path for the v3 scorecard build. Injected v3 into the 7/17-7/23 scorecards the same evening (18:41 .bak batch)."
    },
    {
      "id": "strategy-ruleset-3-0-0",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "property": "strategy_library",
      "title": "Strategy rules ruleset 3.0.0 published",
      "path": "/z/public_html/scorecards/strategy_rules_ruleset-3.0.0.html",
      "completed_at": "2026-07-26",
      "created_at": "2026-07-26",
      "note": "strategy_rules_ruleset-3.0.0.html written 17:04 \u2014 versioned ruleset snapshot backing the v3 scorecard grading."
    },
    {
      "id": "scorecard-v3-history-backfill",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "scorecards",
      "title": "Execution Scorecard v3 backfilled across scorecard history",
      "path": "/z/public_html/scorecards/",
      "completed_at": "2026-07-26",
      "created_at": "2026-07-26",
      "note": "Two waves: 17:10 regenerated 7/17-7/24, then 19:48-19:55 backfilled v3 across the May window (5/01-5/19) and early-July dates. ~30 scorecard_v3_*.html pages now exist alongside their trade_scorecard_* originals, each with a .pre_v3_inject.bak checkpoint."
    },
    {
      "id": "lynch-master-refresh-0726",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "property": "strategy_library",
      "title": "LYNCH_MASTER combinatorial lab refreshed",
      "path": "/z/public_html/system/LYNCH_MASTER.html",
      "completed_at": "2026-07-26",
      "created_at": "2026-07-26",
      "note": "LYNCH_MASTER.html rebuilt 22:00, one week after the original 7/19 build of the 3,672-combo walk-forward lab."
    },
    {
      "id": "v3-live-runner-headless",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "scorecards",
      "title": "v3 live runner wired headless + folded into run_all_eod",
      "path": "/z/setup_daily_reporting/reports/v3_live.sh",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "v3_live.sh + v3_live_hidden.vbs at 16:23 (silent/no-console launch per the headless rule), then run_all_eod.bat updated 17:19 to call the v3 path. v3 is now on the scheduled EOD spine, not a manual build."
    },
    {
      "id": "backside-short-trigger-listener-0727",
      "phase": "phase-alerts",
      "status": "completed",
      "owner": "claude",
      "property": "alerts-hotkeys-etc",
      "title": "Backside Short radar + trigger listener pass",
      "path": "/z/setup_daily_reporting/reports/backside_short.py",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "trigger_listener.py + backside_short.py .bak-0727 (12:31), then backside_short.py + .bak-0727b (12:55). Two checkpointed passes over the one alert path (indicator_alerts -> trigger_listener -> dispatch) and the bear VWAP-fade radar."
    },
    {
      "id": "j3p-swing-radar-page",
      "phase": "phase-j3p",
      "status": "completed",
      "owner": "claude",
      "property": "alerts-hotkeys-etc",
      "title": "J3P Swing Radar surface published",
      "path": "/z/public_html/system/J3P_SWING_RADAR.html",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "J3P_SWING_RADAR.html written 15:40 \u2014 swing/DMA watch surface joining the live J3P stack (WATCHER 08:00, SIP_TODAY 08:45, LIVE 16:15, SETUP_LEAGUE 16:30)."
    },
    {
      "id": "param-surface-monthly-refresh-0727",
      "phase": "phase-param-surface",
      "status": "completed",
      "owner": "claude",
      "property": "param_surface",
      "title": "Param Surface + monthly reports fully regenerated",
      "path": "/z/gt_dev/db_reports/james-monthly-reports/vwap-bands/param-surface/",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "17:08-17:16: all per-symbol param surfaces + heatmaps (ADBE/AMZN/CRWD/HUBS/MSFT/MSTR/MU/SOXL), the combined surface, the full time-of-day set (open/early/mid-am/midday/mid-pm/late) plus tod_comparison + tod_sd_expansion, and the july-2026 / vwap-bands / vwap-research indexes."
    },
    {
      "id": "lynch-archive-page",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "property": "strategy_library",
      "title": "LYNCH archive page published",
      "path": "/z/public_html/system/LYNCH_ARCHIVE.html",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "LYNCH_ARCHIVE.html written 17:00 \u2014 keeps prior LYNCH lab runs reachable now that LYNCH_MASTER is being rebuilt weekly."
    },
    {
      "id": "playbook-data-june-2026-best",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "property": "playbooks",
      "title": "June 2026 playbook data + best-of set rebuilt",
      "path": "/z/setup_daily_reporting/reports/playbook_data/best_june_2026/",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "june_2026_sessions_rth.json + june_2026_trades_raw.json (17:11), then best_june_2026 daily pulls for AAOI / IBIT / NVDA (17:13)."
    },
    {
      "id": "acct-v2-api-v3-campaign-endpoints",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "accountability_hud",
      "title": "Accountability v2: v3 campaign API endpoints (campaigns / campaign / diff)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/v3/",
      "completed_at": "2026-07-24",
      "created_at": "2026-07-24",
      "note": "New api/v3/campaigns.php (22:00), campaign.php (22:02), diff.php (22:02) in the DEV tree \u2014 the campaign-first backend behind Execution Scorecard v3, landed the same night v3 first injected into a published scorecard."
    },
    {
      "id": "acct-v2-hud-drill-ui-0725",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "accountability_hud",
      "title": "Accountability v2 HUD: drill-through UI pass (css + drill.js + chart)",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/hud/",
      "completed_at": "2026-07-25",
      "created_at": "2026-07-25",
      "note": "DEV tree: hud/css/hud.css + hud/js/drill.js (10:56), hud/chart.html (17:20). Front-end half of the v3 campaign drill-through."
    },
    {
      "id": "acct-v2-scorecard-click-chart-0727",
      "phase": "phase-2",
      "status": "completed",
      "owner": "claude",
      "property": "accountability_hud",
      "title": "Accountability v2: scorecard click-through + chart wiring (landed in PROD)",
      "path": "/z/trading/a1p_journal/accountability-v2/hud/js/scorecard-click.js",
      "completed_at": "2026-07-27",
      "created_at": "2026-07-27",
      "note": "PROD tree edited 16:21-17:15: hud/chart.html, hud/js/scorecard-click.js (2 timestamped .bak checkpoints), hud/js/scorecard-chart.js, api/traders.php (+.bak). DEV tree then updated 18:38-18:42 (scorecard-click.js, scorecard-chart.js, hud/index.html). NOTE: source edits in the PROD accountability-v2 tree are unusual \u2014 it normally shows only cache JSON \u2014 so flagging for confirmation that prod and dev are intentionally in sync."
    },
    {
      "id": "sip-tooltip-line-feeds-2026-07-28",
      "phase": "phase-sip",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "SIP + live-scan tooltips use line feeds (and a standing rule for every page)",
      "path": "/z/setup_daily_reporting/reports/build_sip_page.py",
      "completed_at": "2026-07-28",
      "created_at": "2026-07-28",
      "note": "James: tooltips were one long unreadable line. Applied &#10; separators to all title-attribute tooltips across the SIP strips and live watchnames (build_sip_page.py + j3p_live_scan.py, two .bak files kept). Logged as a STANDING RULE in AGENT_SYNC for all agents: every hover/tooltip/title attribute must use line feeds between logical segments, on every current and future page. AGENT_SYNC has the INTENT line but no matching DONE line \u2014 file mtimes confirm the edit landed 07:21."
    },
    {
      "id": "hud-scorecard-chart-mount-guard-2026-07-28",
      "phase": "phase-2",
      "status": "completed",
      "owner": "james",
      "priority": "P2",
      "title": "HUD scorecard-chart: mount-generation guard + ET day bucketing",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/hud/js/scorecard-chart.js",
      "completed_at": "2026-07-28",
      "created_at": "2026-07-28",
      "note": "Dev-tree HUD work (reported only, not edited by the nightly PM). scorecard-chart.js gains _chartMountGen, bumped in dispose() so in-flight fetch callbacks can never touch a removed chart \u2014 the classic teardown race. Also an explicit America/New_York day formatter for bucketing trades per session day, consistent with the ET-everywhere rule. chart.html + index.html updated alongside."
    },
    {
      "id": "vwap-band-research-page-2026-07-28",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "VWAP Band Research System page",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "completed_at": "2026-07-28",
      "created_at": "2026-07-28",
      "note": "Front page for the VWAP Band Entry Surface research lane, published under gt_dev/db_reports/vwap-research/."
    },
    {
      "id": "lynch-forward-test-archive-2026-07-28",
      "phase": "phase-strategy-rules",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "title": "LYNCH Forward-Test Archive published",
      "path": "/z/public_html/system/LYNCH_ARCHIVE.html",
      "completed_at": "2026-07-28",
      "created_at": "2026-07-28",
      "note": "/system/LYNCH_ARCHIVE.html \u2014 keeps the LYNCH scanner's forward-test record so its live signals can be judged against a stored history rather than re-derived each run."
    },
    {
      "id": "nightly-helper-stale-date-fix-2026-07-28",
      "phase": "phase-infra",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "title": "Nightly todos helper rewritten payload-driven (was silently reverting meta to 2026-05-01)",
      "path": "/z/setup_daily_reporting/reports/update_todos_for_nightly.py",
      "completed_at": "2026-07-28",
      "created_at": "2026-07-28",
      "note": "update_todos_for_nightly.py was a hard-coded one-off from the 2026-05-01 run: any bare invocation rewrote meta.updated_at / last_nightly_run / last_nightly_report_* back to 2026-05-01. Tonight's fire hit exactly that. Rewritten to take --payload <json> with meta / tasks / complete sections, plus a read-only --summary mode and an atomic write. It now contains no dates or task data of its own, so it cannot carry a stale date forward."
    },
    {
      "id": "mgp-v2-duplicate-sends-2026-07-28",
      "phase": "phase-mgp",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "BUG: MGP v2 parallel build double-emails every edition",
      "path": "/z/setup_daily_reporting/reports/notify.log",
      "created_at": "2026-07-28",
      "note": "RECURRED 8/03 ON ALL THREE EDITIONS, per notify.log: morning sent 06:49 and again 08:37, midday 11:38 and 11:43, recap 16:41 and 16:43 - six V2 emails where three were expected, to both jdomino@smbcap.com and Jim.Domino@gmail.com. The V1 Domino Briefs sent exactly once each (05:41, 11:31, 16:31), so the duplication is specific to the V2 parallel build. Note that atlas-flow's MGP steps all report PASS, because they check the artifact exists, not that it was emailed once."
    },
    {
      "id": "nightly-scan-default-ssh-a2-2026-07-28",
      "phase": "phase-infra",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "title": "Cover BOTH accountability-v2 trees in the nightly scan (realpath visited-set SHIPPED 8/03)",
      "path": "C:\\Users\\jtdom\\.claude\\scheduled-tasks\\nightly-project-manager\\SKILL.md",
      "created_at": "2026-07-28",
      "note": "HALF SHIPPED 8/03. The realpath visited-set landed in scan_recent_activity.py: scan_recursive now canonicalises every directory with os.path.realpath().casefold() and keeps a seen-set for both directories and files, pruning dirnames in place when a child has already been walked. WebDAV reparse points on the Z: mount made the same physical tree reachable under many literal paths, so the old raw-path check never fired. MEASURED BEFORE: 549 rows for 4 unique files (137 identical copies each) after ~50 minutes. MEASURED AFTER, same root, same date: 14 unique rows in seconds, matching the ssh scan file-for-file. STILL OPEN, which is why this stays pending: half (2), covering the prod HUD tree that the audit hook blocks by name. See also ssh-scan-recursive-public-html-returns-zero-2026-08-02 for the ssh path's own silent-empty failure mode."
    },
    {
      "id": "sip-scoring-rebalance-decision-2026-07-28",
      "phase": "phase-sip",
      "status": "blocked",
      "owner": "james",
      "priority": "P0",
      "title": "DECISION: SIP top-10 is always catalyst-sweep names \u2014 pick a rebalance option",
      "path": "/z/setup_daily_reporting/reports/AGENT_SYNC.md",
      "created_at": "2026-07-28",
      "note": "James asked why the SIP top-10 is always catalyst-sweep names. Verified in the composite: catalyst contributes 21.6-36.0 pts vs mover 1.5, rrg 2.2, liquidity <=10 \u2014 so technical scanner hits (PVP / PBS / GAPMA / LYNCH) effectively cannot reach the top-10 on their own. Three options on James's desk: (1) reserve ~3 of 10 slots for the best non-catalyst technicals, (2) cap catalyst at ~22 with earnings-over-PR tiering, (3) add a scanners component at +6 per family capped at 18. Scoring is blue-line's lane; nothing changes until James picks. BLOCKED ON JAMES."
    },
    {
      "id": "revbreak-long-demotion-decision-2026-07-28",
      "phase": "phase-strategy-rules",
      "status": "blocked",
      "owner": "james",
      "priority": "P0",
      "title": "DECISION: demote revbreak_long from the validated set (confidently negative live)",
      "path": "/z/public_html/system/J3P_SETUP_LEAGUE.html",
      "created_at": "2026-07-28",
      "note": "DAY 6, AND TODAY CUT BOTH WAYS. Tonight's Setup League rebuild (858 resolved signals, 7/07-8/03) puts revbreak_long at n=314, 21% win, -0.31 avg R, LB-90 -0.41, both-halves check failing - the second-worst play on the board and still confidently negative. But it also had a green day: the Robot Report shows 15 trades, 6 winners, +$219. Fifteen trades cannot move a 314-signal verdict, and a green day is precisely the kind of evidence that stalls this call for another week. It has now been live and validated for 6 days past the point the data said to pull it, and it fires again every day it stays in the set."
    },
    {
      "id": "pine-alpha-fl2-session-2026-07-29",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-07-29",
      "title": "Pine session \u2014 Alpha Compression Break + Fashionably Late v2 suite advanced",
      "owner": "James",
      "priority": "P1",
      "property": "strategy-library",
      "note": "Alpha_Compression_Break.pine grew 101,563 -> 110,679 bytes (last prior rev was v165 on 7/21). Same 16:24-16:25 commit cluster also touched SMB2_Fashionably_Late_Unified.pine, Alpha_Setup_Playbook.pine, Alpha_Trend_Alignment_Unified.pine, _patch_gates.py and smb_fl2_kore_strategy.py, plus the .cursor pine-script-assignments rule. J3P_VCET.txt and James-ThreePlayv2.6.2.txt were touched earlier the same day.",
      "path": "/z/trading/dev/a1p_journal/cursor/TradingView/dev/"
    },
    {
      "id": "sentiment-accuracy-pages-2026-07-31",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-07-31",
      "title": "Sentiment-accuracy scoring pages published (SENTIMENT_ACCURACY + MTF_SENTIMENT_ACCURACY)",
      "owner": "James",
      "priority": "P1",
      "property": "system-hub",
      "note": "/system/SENTIMENT_ACCURACY.html (25,109 b) and /system/MTF_SENTIMENT_ACCURACY.html (9,826 b) both first appear on 7/31 (08:53 / 08:38 ET) and are absent from every prior nightly scan. These close the loop on whether the MTF/sentiment read is actually right, which is the prerequisite for trusting it as a selectivity gate.",
      "path": "/system/SENTIMENT_ACCURACY.html"
    },
    {
      "id": "j3p-swing-radar-cup-verify-2026-07-31",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-07-31",
      "title": "J3P surfaces published \u2014 Swing Radar, Cup Verify, Watcher",
      "owner": "James",
      "priority": "P2",
      "property": "j3p",
      "note": "/system/J3P_SWING_RADAR.html (15:40), /system/J3P_CUP_VERIFY.html + .json (20:26) and /system/J3P_WATCHER.html (08:01) all first appear 7/31. Cup Verify follows the existing PLAY_VERIFY pattern (html + json pair) so the pattern is now applied to a second setup family.",
      "path": "/system/J3P_SWING_RADAR.html"
    },
    {
      "id": "eod-catchup-trader-backfill-2026-07-31",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-07-31",
      "title": "EOD catch-up backfilled 18 trader-days clean (ok=18 fail=0)",
      "owner": "James",
      "priority": "P1",
      "property": "accountability",
      "note": "eod_catchup.log 2026-07-31T16:45:02: pulled 7/29-7/31 for nolan (14860), brian (14959) and thomas (14960) via kti trades date, 18/18 succeeded, and confirmed 'scorecard present now: True'. Caveat worth chasing separately: every fetched day logged 'DB enrichment: 0 entries with regime/playbook tags', so the fills landed untagged \u2014 strategy attribution on those days will be blank.",
      "path": "/z/setup_daily_reporting/reports/eod_catchup.log"
    },
    {
      "id": "home-machine-sleep-burst-2026-08-01",
      "phase": "design",
      "status": "pending",
      "title": "HOME machine sleeping through its schedule \u2014 whole task backlog burst-fires on wake",
      "owner": "James",
      "priority": "P0",
      "property": "infrastructure",
      "note": "Sat 2026-08-01: DESKTOP-AKU8I62 woke at 14:12 ET and fired everything at once \u2014 nightly-PM heartbeat 14:12:13, Mid-Day brief 14:12:46, Morning Game Plan 14:13:21, EOD Recap 14:15:52. A Morning Game Plan delivered at 2:13 PM is worthless to a trader. Same wake also explains multi_tf_daemon.log stopping dead at 7/31 20:00 with no 8/01 cycles. a2 cron is NOT the problem (it is timeout-guarded and weekday-only and ran clean); this is Windows-side sleep. Fix candidates: powercfg wake timers / 'run task as soon as possible after a missed start' being the thing that causes the burst rather than preventing it. CORRECTION 8/1 23:06: the multi-tf daemon is NOT evidence for this bug. It logged a clean 'daemon exiting' at 2026-07-31 20:00:00 on its normal 04:00-20:00 schedule and takes no Saturday start (7/25 likewise has none). Remove it from the symptom list. What DOES stand: nightly produced zero LAUNCHED lines on 7/29, 7/30 and 7/31, then burst-fired everything on wake at 14:12 on 8/1, delivering a Morning Game Plan at 2:13 PM Saturday. Tonight's 23:06 fire was on time - first since 7/28 - so the box stayed awake through Saturday evening at least.",
      "path": "/z/logs/nightly-startup.log"
    },
    {
      "id": "multi-tf-daemon-runaway-2026-07-31",
      "phase": "design",
      "status": "pending",
      "title": "multi-tf daemon rewrites an 800 KB page every ~60s and keeps going hours past the close",
      "owner": "James",
      "priority": "P1",
      "property": "mtf-vp",
      "note": "multi_tf_daemon.log shows full rebuild cycles #859-#869 running 19:49-20:00 ET on 7/31 \u2014 four hours after the cash close \u2014 each one rewriting /system/multi-tf-stack/<date>.html at ~800-810 KB over WebDAV. Byte size is identical across most consecutive cycles, i.e. the rebuild is a no-op. The run_if_stale.sh pattern already used by PVP is the obvious fix, plus an RTH window guard.",
      "path": "/z/setup_daily_reporting/reports/multi_tf_daemon.log"
    },
    {
      "id": "mgp-weekend-edition-suppress-2026-08-01",
      "phase": "design",
      "status": "pending",
      "title": "Suppress or relabel the weekend brief editions",
      "owner": "James",
      "priority": "P2",
      "property": "mgp",
      "note": "All three editions generated and emailed on Saturday 8/01. To its credit the body is honest \u2014 it says 'Saturday post-close recap' and stamps every number as-of 2026-07-31 \u2014 but the subject line reads 'End of Day Recap 2026-08-01 - chop day' and the tape read says 'Session done: no clean directional edge today', which describes a session that never happened. Either skip Sat/Sun or relabel them as a Friday carry-forward.",
      "path": "/z/public_html/mgp/2026-08-01-recap.html"
    },
    {
      "id": "notify-log-truncation-2026-07-31",
      "phase": "design",
      "status": "pending",
      "title": "notify.log was reset \u2014 send history before 7/31 16:30 is gone",
      "owner": "James",
      "priority": "P2",
      "property": "mgp",
      "note": "notify.log is down to 14 lines, earliest entry 2026-07-31T16:30:54. That is the audit trail used to prove which briefs actually emailed, so its loss is why the MGP duplicate-send bug could only be partially checked this run. Needs rotation that archives rather than truncates.",
      "path": "/z/setup_daily_reporting/reports/notify.log"
    },
    {
      "id": "nightly-heartbeat-path-mangling-2026-08-01",
      "phase": "build",
      "status": "pending",
      "title": "BUG: heartbeat DONE line records report=C:/Program Files/Git/todos/... instead of /todos/...",
      "owner": "Claude",
      "priority": "P3",
      "note": "Git Bash MSYS path translation rewrites the leading-slash argument, so nightly-startup.log stores 'report=C:/Program Files/Git/todos/daily/<date>.html' for 7/15, 7/18, 7/19, 7/28 and the 8/1 14:12 run. 7/27 is correct because it was passed without a leading slash. Cosmetic today, but it makes the log's report links useless for the morning diagnostic check. Fix: pass the path without a leading slash, or set MSYS_NO_PATHCONV=1 for that call.",
      "path": "Z:\\setup_daily_reporting\\reports\\heartbeat.py",
      "created_at": "2026-08-01",
      "property": "todo"
    },
    {
      "id": "lynch-master-combinatorial-2026-08-02",
      "phase": "validate",
      "status": "completed",
      "completed_at": "2026-08-02",
      "title": "LYNCH Master Combinatorial Backtest published - 3,618 combinations, walk-forward validated",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "path": "/system/LYNCH_MASTER.html",
      "note": "3,618 Grade x Sizing x Exit x Regime-gate combinations, 5 years, real dollars, whole shares, $50k position cap, SMB commissions. 998 survived the both-halves test. Winner on BOTH return/DD and raw dollars: A+B+C / UVXY-mom x Reactive / Swing pure hold-5d / no gate -> +$71,454, $14,992 max DD, 477% return/DD, 45.1% win, Sharpe 0.74. Top-50 concordance says grade (100%) and sizing (84%) are the real edge; exit tops out at 26% and gate splits 36/36/28, so both are close to noise."
    },
    {
      "id": "boat-v46-filter-cost-sweep-2026-08-02",
      "phase": "validate",
      "status": "completed",
      "completed_at": "2026-08-02",
      "title": "BOAT v4.6 filter + cost sweep benched on a2 (efficiency gate, OR%, per-symbol-day cap, TZ fix)",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "path": "a2:~/j3p-v269/ (boat_bench.py, boat_bench2.py, bench_frozen.py, bench_tz.py, target_sweep.py; runs/filters.log, liquid.log, tzfix.log, boat_v46.json)",
      "note": "139 dates, top 8 gappers/day. Signal is real - beats random control at every setting (-0.042 vs -0.191 capped; -0.070 vs -0.130 uncapped) - but does not clear costs: avgR +0.006 at 0.00%, -0.018 at 0.05%, -0.042 at 0.10%, -0.066 at 0.15%. Selectivity is the differentiator: cap=1 symbol-day gives 1.0 trades/day and IMPROVES out of sample (2nd half -0.032 vs 1st -0.052); uncapped in-play-only gives 7.6/day and DECAYS (2nd half -0.094 vs 1st -0.041). kind=BOAT loses to kind=V in every cut."
    },
    {
      "id": "boat-v46-cost-hurdle-decision-2026-08-02",
      "phase": "decide",
      "status": "pending",
      "created_at": "2026-08-02",
      "title": "DECISION: BOAT v4.6 - commit to the cap=1 selectivity build and attack costs, or shelve it",
      "owner": "James",
      "priority": "P1",
      "property": "strategy_library",
      "path": "/system/ (bench: a2:~/j3p-v269/runs/)",
      "note": "THE DATA ANSWERED THIS ON 8/03 - see boat-v46-data-faults-falsify-2026-08-03. Option (a), lock cap=1 and attack the cost side, rested on the 8/02 finding that cap=1 got BETTER out of sample. That finding is dead: it was an artifact of stops floored at nothing, one of which was 0.58 cents. With the stop floored at 0.5 ATR and the EDT timezone hard-code fixed, the strategy is negative in both halves at every setting swept, and the best surviving build is avgR -0.030 at 0.10% cost with t=-1.0. There is no cost-side story left to attack. Recommendation is now unambiguous: option (b), shelve BOAT and move the bench hours to the queued depth-kill / band-stop / 3-way-exit work. Left pending because it is James's call to make, not an automatic close."
    },
    {
      "id": "multi-tf-verdict-preview-timeout-chronic-2026-08-02",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-02",
      "title": "BUG: multi-tf daemon 'verdict preview timed out (>60s)' fires every single day since at least 7/20",
      "owner": "Claude",
      "priority": "P2",
      "property": "mtf_vp_scanner",
      "path": "setup_daily_reporting/reports/multi_tf_daemon.log, gen_multi_tf_stack.py",
      "note": "WORST DAY YET ON 8/03: 39 'verdict preview timed out (>60s)' lines, against a prior daily range of 2-22 across 7/20-8/02. The daemon itself was healthy - 869 full plus 1594 tick refreshes, exiting cleanly at 20:00 ET - and the preview does land at exit, so the page is not stale. Separately, this closes the 8/03 EOD brief's worry that the daemon log had no 8/03 entries as of 18:16 ET: the log is buffered until exit, and it holds 886 lines for today. So the log is trustworthy; the timeout is real and getting worse."
    },
    {
      "id": "ssh-scan-recursive-public-html-returns-zero-2026-08-02",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-02",
      "title": "Nightly scan: heavy `find` over public_html on a2 is killed host-side (exit 255 / silent empty)",
      "owner": "Claude",
      "priority": "P2",
      "property": "pm_system",
      "path": "scheduled-tasks/nightly-project-manager/SKILL.md",
      "note": "CONFIRMED IN PRACTICE 8/03. Tonight's scan was run entirely by the three-part discipline from this ticket - narrow roots (one or two per ssh call), no pipe on the remote side, empty treated as suspicious - and every call returned honest non-empty results in seconds, including the ~/public_html/system + ~/public_html/scorecards pair that the wide walk could not do. The 8/03 EOD brief independently reproduced the failure: a single wide scan of ~/public_html returned 0 rows with exit 0 while narrow scans of that same tree returned 23 files. So the workaround is proven; what is still missing is codifying it in the SKILL so a future fire cannot revert to a wide walk and read the silent empty as a quiet day."
    },
    {
      "id": "hud-error-log-db-config-spam-2026-08-02",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-02",
      "title": "Housekeeping: per-minute poller writes DB name + DB user into web-tree error_log files",
      "owner": "Claude",
      "priority": "P3",
      "property": "accountability_hud",
      "path": "trading/dev/a1p_journal/accountability-v2/api/error_log, gt_dev/php_error_analysis.log",
      "note": "Both files grow every 60 seconds with 'DB Config / Database: trsports_tradingjournal_dev / DB User: trsports_trader / Host: localhost' followed by 'Database connection established successfully.' These are not errors - it is success logging in a file named error_log, inside the web tree. Two problems: it buries real errors, and it prints the database name and user into a file that may be web-readable. Drop the success branch to a debug flag and confirm the path is not served."
    },
    {
      "id": "boat-v46-data-faults-falsify-2026-08-03",
      "phase": "validate",
      "status": "completed",
      "completed_at": "2026-08-03",
      "created_at": "2026-08-03",
      "title": "BOAT v4.6 falsified - EDT timezone hard-code + no-minimum-stop faults found, fixed, verdict flipped negative in both halves",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "path": "a2:~/j3p-v269/ (bench_v2.py, bench_v3.py, bench_v4.py, boat_bench2.py, export_boat_trades.py, runs/minstop.log, runs/atrp.log, runs/brk.log, runs/combo.log) + /system/J3P_BOAT_TRADES.html",
      "note": "Two data faults found, both of which made every earlier BOAT result look better than it was. (1) TIMEZONE: the session window was hard-coded to EDT, so roughly a third of the 139-date sample read 08:30-15:00 instead of 09:30-16:00 - it traded thin pre-market bars as if they were the open and threw away the closing hour. (2) STOP FLOOR: the stop had no minimum distance, and the stop distance is the denominator of every R. Some stops were a fraction of a cent; one trade scored +285R on a 0.58 cent stop. The stop is now floored at 0.5 ATR. WITH BOTH FIXED THE STRATEGY IS NEGATIVE IN BOTH HALVES at every setting swept: ATR floor 0/0.15/0.25/0.40/0.60% of price, neckline buffer 0/0.25/0.50/0.75/1.00 ATR, min stop 0.3 ATR. Representative: n=6567, avgR -0.075, win 48.6%, t=-5.3, first half -0.047 / second half -0.099. The 8/02 finding that cap=1 improved out of sample was an artifact of the tiny stops. Best surviving build - ATR>=0.6% + chop filter + min stop + one trade per symbol-day - is n=1022, avgR -0.030 at 0.10% round trip, 52.7% win, t=-1.0, halves -0.040 / -0.020; at 0.00% cost it is +0.009. It still beats its random control everywhere (-0.149 for the same n), so the signal is real - it just does not clear a round trip. The 10-best / 10-worst execution page was rebuilt on the corrected numbers and leads with the two faults."
    },
    {
      "id": "atlas-flow-16-step-verifier-green-2026-08-03",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-03",
      "created_at": "2026-08-03",
      "title": "Atlas Flow expanded to a 16-step end-to-end pipeline verifier - GREEN 16/16, clears the V1-morning FAIL that was RED since 7/31",
      "owner": "Claude",
      "priority": "P1",
      "property": "pm_system",
      "path": "/system/atlas-flow-status.html + atlas-flow-full.html (atlas_flow_test.py on a2)",
      "note": "Ran 19:00:02 ET, 16 PASS / 0 FAIL / 0 SKIP in 5.6s on a2. Coverage is now the whole chain, each step with its own pre-check, exec line and post-check: pre-flight (DB 6853 rows, reports dir, all 4 scanners fresh); 3 scanners (PVP 120 picks age 96s, PBS 110 alerts, GAPMA 1190 longs + 516 shorts); 4 SIP steps (update_scanners 250 rows / 250 unique tickers, run_sip_daily 10 composite picks topped by AXTI/SPCX/ULTA, build_sip_page 627 KB with 10 pick-readiness divs, update_daily_stats with all top 10 carrying a bias); 2 multi-TF (index 802 KB age 41s, sidecar OK 5 panels, session_verdict age 28s); and all 6 MGP editions V1/V2 x morning/midday/recap. This cleared the V1-morning FAIL that had held the monitor RED since Friday 7/31. ONE WEAK CHECK TO TIGHTEN: PBS reported age 10480s (2h 55m) and still passed - the step has no freshness threshold, so a stale scanner reads as healthy."
    },
    {
      "id": "j3p-robot-report-2026-08-03",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-03",
      "created_at": "2026-08-03",
      "title": "J3P Robot Report - 38 simulated trades across 5 gate variants, robot +$2,509 vs live account +$24.11",
      "owner": "Claude",
      "priority": "P1",
      "property": "j3p",
      "path": "/system/J3P_ROBOT_REPORT.html",
      "note": "Built 16:17 ET. Takes every signal the scanner fired, follows each play's own exit rules at $100 risk per trade, and closes the day at +$2,509 on 38 trades without ever coming near the $500 daily stop - roughly 100x the live account's +$24.11 net. Per play: vwap_reject_short 5 trades / 4 winners / +$1,330 (LB-90 +0.13 on n=232, still the only green lower bound on the board and best again today); earn_gap_reclaim 4/4 for +$545 (LB-90 -0.14, n=69, LAB, perfect day but under-sampled); vwap_pullin 5 of 9 for +$533 (LB-90 -0.15, n=150, LAB); revbreak_long 6 of 15 for +$219 (LB-90 -0.41, n=314, green day but confirmed bleeder); trend_ride 2 of 5 for -$118 (LB-90 -0.53, n=66, worst lower bound on the board). The value is that the robot and the Setup League disagree today, and the page shows both scoreboards side by side instead of picking one."
    },
    {
      "id": "setup-league-rebuild-858-signals-2026-08-03",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-03",
      "created_at": "2026-08-03",
      "title": "Setup League rebuilt on 858 resolved signals (2026-07-07 to 2026-08-03)",
      "owner": "Claude",
      "priority": "P2",
      "property": "strategy_library",
      "path": "/system/J3P_SETUP_LEAGUE.html",
      "note": "Rebuilt 16:30 ET. This is the scoreboard the two open P0 decisions hang on. vwap_reject_short remains the only play with a green 90% lower bound (+0.13, n=232). revbreak_long lands at n=314, 21% win, -0.31 avg R, LB-90 -0.41, both-halves check failing. trend_ride is the worst lower bound at -0.53 on n=66."
    },
    {
      "id": "sprint-status-brief-on-time-2026-08-03",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-03",
      "created_at": "2026-08-03",
      "title": "First on-time sprint-status brief in six weeks - published 07:23 ET, 67 minutes before planning",
      "owner": "Claude",
      "priority": "P2",
      "property": "pm_system",
      "path": "/system/briefings/2026-08-03-sprint-status.html",
      "note": "Scheduled 07:00, published 07:23, 67 minutes before the 08:30 planning meeting - a real agenda instead of a retrospective. It is trustworthy for the first time because nightly-helper-stale-date-fix-2026-07-28 stopped the writer silently reverting meta to 2026-05-01: last Monday the same query reported 0 closes for 7/20-7/24, today it returns 9. Headline numbers: 16 tickets closed 7/27-7/31, the best week in the 4-week series (7 -> 11 -> 9 -> 16), but ZERO of them were Sprint 5 committed scope - all 16 were unplanned work. 39 open P0, 75 open P1, 15 carry-over from Sprints 1/3/4, board still 39 days behind the calendar with active_sprint_id = 1."
    },
    {
      "id": "plan-adherence-1-of-249-2026-08-03",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-03",
      "title": "Plan adherence collapsed to 1 of 249 picks (99.6% off book) on 8/03",
      "owner": "James",
      "priority": "P1",
      "property": "accountability_hud",
      "path": "/scorecards/trade_scorecard_2026-08-03.html",
      "note": "Worse than the 93% off-book baseline in the Execution Report. Account 14859 finished +$39.71 gross / +$24.11 net on 181 round trips and 265 fills for 780 shares across 6 names, 48.5% win, B grade - a green day, but commission ate $18.88 of $39.73 gross (47.5%) and TSLA alone took 119 fills, 3.2x the median name. Exactly one A-grade pick was traded (ORCL, -$34.89). In fairness, two untraded picks would have stopped out, so the skips were right - but the day was almost entirely unplanned, and the Robot Report shows the same signals returning +$2,509 when followed mechanically. Raised by the 8/03 EOD brief with no ticket behind it; ticketed here."
    },
    {
      "id": "catchup-zero-regime-playbook-tags-2026-08-03",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-03",
      "title": "EOD catch-up reports '0 entries with regime/playbook tags' on every trader-day",
      "owner": "Claude",
      "priority": "P1",
      "property": "accountability_hud",
      "path": "setup_daily_reporting/reports/eod_catchup.log",
      "note": "The 16:45 ET catch-up ran clean - ok=18 fail=0 across all 6 traders for 7/30, 7/31 and 8/03 - but every single trader-day that returned fills also logged 'DB enrichment: 0 entries with regime/playbook tags'. The classifier is not writing into the scorecard path, so the scorecards are outcome-only with no strategy attribution. This is downstream of the scientific_layer migrations 001-009 (run-scientific-layer-migrations, 104 days open), which is the argument for running them: the enrichment cannot work until that path exists. Confirm the dependency before building anything new here."
    },
    {
      "id": "atlas-flow-pbs-staleness-threshold-2026-08-03",
      "phase": "fix",
      "status": "pending",
      "created_at": "2026-08-03",
      "title": "Atlas Flow scanner steps have no freshness threshold - PBS passed at 2h 55m stale",
      "owner": "Claude",
      "priority": "P3",
      "property": "pm_system",
      "path": "/system/atlas-flow-status.json (atlas_flow_test.py on a2)",
      "note": "In tonight's GREEN 16/16 run, step 0.5b PBS pullback reported '110 alerts, age 10480s' - just under three hours old - and still returned PASS with no post-check. PVP by contrast was 96s old. A monitor that reads a nearly-three-hour-old scanner as healthy is the same class of blind spot the freshness dashboard exists to close. Add a per-scanner max-age and fail past it."
    },
    {
      "id": "stockbee20-breadth-pine-2026-08-04",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-04",
      "title": "Stockbee 20% breadth: Pine up/down scanner pair + HOW_TO wiring TV to indicator_alerts",
      "owner": "claude",
      "priority": "P2",
      "property": "strategies",
      "path": "/z/public_html/system/stockbee20/",
      "note": "stockbee20_up.pine + stockbee20_down.pine + HOW_TO.md. Two TV alerts post to the existing SMB webhook; aggregator counts today's fires split by tape type; surfaces on the multi-tf-stack cheat card, Cross-Signal Tensions, and all three MGP editions. Paste-and-go setup doc written for a 5-minute install."
    },
    {
      "id": "sip-daily-csv-export-2026-08-04",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-04",
      "title": "Publish the daily SIP as CSV (latest + permanent per-day archive)",
      "owner": "claude",
      "priority": "P2",
      "property": "sip",
      "path": "/z/setup_daily_reporting/reports/sip_page/j3p_sip_today_page.py",
      "note": "Written on every SIP build (7:00 + 8:45 + the 15-min watchdog): /sip/sip_today.csv always latest, /sip/csv/sip_YYYY-MM-DD.csv permanent. Columns: date,rank,symbol,score,reason,scanners(setup_names),earnings_tier,earnings_badge,catalyst_news. py_compile clean on a2 3.9. Logged in AGENT_SYNC by the earnings agent."
    },
    {
      "id": "amsb-playbook-crwd-2026-08-04",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-08-04",
      "title": "Adaptive Mean + sigma Bands v5.2 TOOL PLAYBOOK + CRWD case + v6 Pine",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/public_html/playbooks/adaptive-mean-sigma-bands.html",
      "note": "8-slide playbook: one engine / three zooms (daily, 30-min, 2-min), 9 tracked trade types, CRWD 8/4 close 211.22 (+6.63R off 191.21, projected 209.15 tagged and paid, sitting at +1.93 sigma). Carries a STAT-HONESTY correction of the deck: '26% win' was not a win rate and the 168 signals were not this exact setup - paste-ready slide fixes included. Banner says HYPOTHESIS, not validated $-edge. Tool source at playbooks/_tools/adaptive_mean_swing_v6.pine; index.html got a tool card (backup index.html.bak-pre-tool-card-2026-08-04_2243)."
    },
    {
      "id": "amsb-alert-json-plumbing-2026-08-04",
      "phase": "build",
      "status": "pending",
      "title": "Wire AMSB into indicator_alerts: replace fixed alertcondition() with alert() JSON payloads",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "created_at": "2026-08-04",
      "path": "/z/public_html/playbooks/_tools/adaptive_mean_swing_v6.pine",
      "note": "The indicator's 16 alertcondition() lines send a fixed sentence only - no price, stop, target, size or bar time - so no AMSB firing can ever reach Postgres. Fix: on the entry bar fire alert('AMSB|{...}') with indicator/sym/tf/event_type=ENTRY/direction/entry_px/stop_px/tgt_px/qty/setup/ts; same shape on exit with event_type=EXIT, exit_px, reason (STOP/TARGET/WAVE5/TIME) and R; carry a trade_id on both legs so they pair. Receiver requires only sym/tf/event_type/indicator and keeps extras in raw_payload. Path: tradingview_smb_indicator_trigger.php -> alerts_tradingview.indicator_alerts -> scorecards + trade-check + playbook builder. This is PLUMBING, so the test-before-ship rule does not gate it; any entry/exit LOGIC change still goes through the a2 engine first. Spec written out in the playbook's 'Capturing the next one' slide."
    },
    {
      "id": "amsb-hold-rules-backtest-2026-08-04",
      "phase": "research",
      "status": "pending",
      "title": "Backtest the three AMSB hold rules (the tool exits at target and says nothing about holding)",
      "owner": "claude",
      "priority": "P2",
      "property": "playbooks",
      "created_at": "2026-08-04",
      "path": "/z/public_html/playbooks/adaptive-mean-sigma-bands.html",
      "note": "Blind spot named in the playbook: AMSB exits at its own target, so its +0.48R-style numbers have nothing to say about 'I should have held for 6R'. Three candidate hold rules are written up and untested. Do NOT put the tool's expectancy next to a hold claim until these are run in the a2 bench. Same class of error as the '26% win' misread the playbook corrects."
    },
    {
      "id": "snap-gap-minus-100-guard-2026-08-04",
      "phase": "fix",
      "status": "pending",
      "title": "Guard the SIP composite gap calc - SNAP emitted gap:-100.0%",
      "owner": "claude",
      "priority": "P2",
      "property": "sip",
      "created_at": "2026-08-04",
      "path": "/z/setup_daily_reporting/reports/enrich_picks.py",
      "note": "Flagged in AGENT_SYNC 2026-08-04 by the earnings agent, addressed to blue-line's scoring lane. Today's composite reason for SNAP read 'gap:-100.0%' - almost certainly a missing/zero prior close rather than a real move. It shipped into the new /sip/csv/sip_2026-08-04.csv archive, so the bad value is now permanent in a file traders read. Wanted: sanity bound on the gap term plus a visible NULL rather than a fabricated -100%."
    },
    {
      "id": "hud-dev-edits-2026-08-04",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-04",
      "title": "HUD dev-tree edits: api/trader.php + hud/js/render.js + hud/index.html",
      "owner": "james",
      "priority": "P2",
      "property": "accountability",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/",
      "note": "BOOKKEEPING ONLY - recorded so the day is not blank on the calendar. Evidence is file mtimes alone: three source files edited 18:17-18:26 ET in the DEV accountability-v2 tree (api/trader.php 18:17:47, hud/js/render.js 18:23:22, hud/index.html 18:26:41). The nightly PM did not open the files, so WHAT changed is not established here - do not read this entry as a shipped feature. Cockpit is read-only to the nightly run per feedback_hud_off_limits (reporting is allowed, editing is not). Also noted: accountability-v2/api/error_log was being written at 23:08 ET and could not be opened over WebDAV during this run - worth a look if the HUD misbehaves."
    },
    {
      "id": "swing-playbook-engine-port-2026-08-05",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-05",
      "title": "Swing Playbook Engine: faithful v5.2 Python port + CRWD validation gate PASSED",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/",
      "note": "Tasks 1-2 of the Swing Playbook Engine work order. NEW tree engine.py / fetch_bars.py / validate_crwd.py / data/ / out/ / DECISIONS.md, version v5.2-port-1.0.0. Gate on CRWD 2026-07-31 30-min RTH: state_card_reconciled true AND trade_levels_reconciled true - setup bar 12:30, entry bar 13:00, basis 182.51, bw pctile 13.5, COMPRESSION-BASE, entry 191.21 / stop 187.98 / target 197.68 all matched to the penny. Massive adjusted 1-min resampled to 30m RTH; no shared files touched, no DB writes."
    },
    {
      "id": "swing-engine-stop-rule-ruling-2026-08-05",
      "phase": "design",
      "status": "completed",
      "completed_at": "2026-08-05",
      "title": "Ruling closed: Play 1 stop rule becomes a tournament axis, not a fix",
      "owner": "james",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/DECISIONS.md",
      "note": "The v5.2 .pine literal structStop = min(hold low, lowest(low,3)) gives 187.18 / risk 4.03 / 2R 199.27; the playbook + work-order reference use the hold low alone, 187.975 / risk 3.235 / 2R 197.68. 24.6% risk delta, which rescales every R downstream. James ruled: do NOT pick one - carry both as a fifth member of the Task-5 STOP GRID next to {hold low / 1.3-sigma band / session LOD / 30m 5-DMA} and let the data decide. Implemented as Params.vwap_stop_mode. Logged append-only in DECISIONS.md."
    },
    {
      "id": "swing-census-dual-stop-columns-2026-08-05",
      "phase": "build",
      "status": "completed",
      "title": "Census must emit BOTH stop candidates per signal (stop_pine_literal + stop_hold_low)",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "created_at": "2026-08-05",
      "path": "/z/trading/dev/swing-playbook-engine/engine.py",
      "note": "Direct consequence of today's stop-rule ruling. If Task 3's census writes a single 'stop' column, the Task-5 tournament cannot score the two stop modes without a full re-scan of the universe. Write stop_pine_literal AND stop_hold_low to swing_signals from the first census run. Cheap now, expensive later.",
      "completed_at": "2026-08-06"
    },
    {
      "id": "swing-engine-leg-filter-tv-export-2026-08-05",
      "phase": "fix",
      "status": "completed",
      "title": "BLOCKED ON JAMES: TradingView settings export - leg filter misses the reference signal by 1.6%",
      "owner": "james",
      "priority": "P0",
      "property": "playbooks",
      "created_at": "2026-08-05",
      "path": "/z/trading/dev/swing-playbook-engine/DECISIONS.md",
      "note": "Arming requires the ride to travel >= vpLeg (1.0) x ATR from session VWAP. At the 7/31 12:30 setup bar the port measures leg 2.2455 vs ATR 2.2821 - short by 0.0366, i.e. 1.60% of ATR. Every level and state-card field reconciles once the signal is forced, but the engine does not fire it unaided, so the census cannot be trusted to find these trades. Three candidates: (a) James's saved chart settings hold a vpLeg below 1.0, (b) TradingView's consolidated feed prints the 12:30 high a few cents above Massive's 189.80, (c) ATR parity. Needs the TradingView settings export to resolve. Fragility note from the port: 18 of 73 leg tests in the 1131-bar window land within 15% of the threshold, so this filter is twitchy either way and should be carried as a census sensitivity, not treated as exact.",
      "completed_at": "2026-08-06"
    },
    {
      "id": "crwd-day1-pullback-playbook-2026-08-05",
      "phase": "ship",
      "status": "completed",
      "completed_at": "2026-08-05",
      "title": "CRWD 7/31 case-study playbook published: Day 1 Pullback to the 50 DMA",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/public_html/playbooks/crwd-2026-07-31.html",
      "note": "Full 12-section case study of the trade the engine was built to reproduce: bigger picture, the setup (pullback / base / reclaim), daily technical, the 30-min trigger frame, annotated 7/31 + 8/3 replay, management, reading the tape, the Adaptive Mean + sigma Bands technology section, the NVDA -> CRWD read, and a scorecard + DATA TRUST closer. Entry 191.21, risk 3.23 / 2.00R, +2.00R banked at 197.68, 8/4 close 211.22, daily +2 sigma 211.87 no-adds and +3 sigma 221.31 trim objective. Playbooks index card added (backup index.html.bak-pre-crwd-card-2026-08-05_0610)."
    },
    {
      "id": "vwap-band-research-index-2026-08-05",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-05",
      "title": "VWAP Band Research System index page",
      "owner": "claude",
      "priority": "P2",
      "property": "param_surface",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "note": "Front door for the VWAP band research lane written 16:46 ET. Bookkeeping entry from the file scan - the page exists and is titled 'VWAP Band Research System'; the nightly run did not audit its contents, so do not read this as a validated result set."
    },
    {
      "id": "swing-leg-filter-gate-closed-2026-08-06",
      "phase": "fix",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "Swing engine gate CLOSED: v5.2 as written does not fire the reference trade - port is correct",
      "owner": "claude",
      "priority": "P0",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/DECISIONS.md",
      "note": "Resolved WITHOUT the TradingView export. Verified against the source file: vpLeg default IS 1.0 (line 380), leg test line 1013, structStop min() line 1218. No saved Play-1 settings exist anywhere (the live chart has no Play-1 inputs), so vpLeg=1.0 stands and the leg test misses the 7/31 CRWD setup by 1.6% of ATR. Conclusion: the port is faithful; v5.2's Play 1 is marginally stricter than James traded. vpLeg becomes a TUNING AXIS in Task 5 rather than a fixed number - 18 of 73 leg tests sit within 15% of the threshold, so that one input silently decides ~25% of Play 1 signals. Also resolved same day: Basis MA type = SMA (spec default + the 7/31 '5DMA flat' state card both confirm; the EMA setting belongs to the older pre-v3.4 live script). Tasks 1-2 complete, Task 3 unblocked."
    },
    {
      "id": "swing-census-v2-scaled-2026-08-06",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "Census scaled 30 -> 150 names: 21 signals rejected as a tournament base, rebuilt to 344",
      "owner": "claude",
      "priority": "P0",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/out/census_summary.json",
      "note": "Census v1 (30 names / 18 months) returned 21 Play-1 signals - too few for a 216-combo hold-rule grid x 5 stops. James's ruling: scale the SAMPLE, do not loosen the RULE (dropping the PDH gate or lowering vpLeg would change the strategy under test and manufacture the n). New frozen universe: top 150 common stock by dollar volume ranked 2022-12-30, bars from 2022-10-01; CRWD qualifies on its own merit at that rank date. census-1.0.0 funnel: A daily-qualified 11,969 symbol-days -> B 30m coil 41,904 (3,920 coil-days) -> C PDH break 949 -> D Play 1 fired 344. New scripts build_universe.py, fetch_census.py, census.py."
    },
    {
      "id": "swing-census-atr-floor-bug-2026-08-06",
      "phase": "fix",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "Bug found + fixed: census stored RAW structural stops without the engine's ATR floor",
      "owner": "claude",
      "priority": "P0",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/census.py",
      "note": "Census v1 wrote structural stops without applying min(structStop, close - 0.5 x ATR). One signal carried an 8-cent stop on a $234 AMZN, which inflated every R derived from it. The floor is now applied to all five stop candidates and it widened the stop on 5 of 21 v1 signals. Raw values retained alongside with a _raw suffix so the correction is auditable. Self-caught, not reported."
    },
    {
      "id": "swing-hold-rule-tournament-v1-2026-08-06",
      "phase": "research",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "Hold-rule tournament run: 810 combos (162 hold-rule x 5 stop) over 344 signals",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/out/tournament.log",
      "note": "Axes: bank (0/0.333/0.5) x gate (always / vwap_only / break+vwap) x night-1 stop (hold_low / day_lod / sigma_1_3) x trail (4) x mm_scale x 5-member stop grid. Best ExpR +1.125R (bank 0, vwap_only, hold_low night-1, daily_5dma_close trail, stop_session_lod, n=86, t=1.57). The Day-1 Hold Rule v1 benchmark ranks 667 of 810 at +0.073R - the current live rule is near the bottom. Worst combos all pair rail_50min trail with stop_pine_literal (negative ExpR at n=344). 84 of 810 combos pass both-halves + n>=30 + t>=2. Chop leak by stop: hold_low 47% of stopped trades printed 2R within 26 bars, 5dma_30m 47%, pine_literal 45%, sigma_1_3 34%, session_lod 25%. READ AS PRELIMINARY - built on static stops, see the migrating-stop task."
    },
    {
      "id": "swing-stop-sweep-2026-08-06",
      "phase": "research",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "ATR stop sweep: no static stop level beats holding with no stop at all",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "path": "/z/trading/dev/swing-playbook-engine/out/stop_sweep.csv",
      "note": "0.25x to 4.0x ATR plus a no-stop control, 344 signals, 10- and 20-session horizons. At 20 sessions the no-stop control returns mean +2.63 ATR, 56.4% win, t=3.16, both halves positive (+1.46 / +3.80), top-5 share 26%. Every stop level scores below it and stop-out rates run 76-97%. At 10 sessions nothing is significant either way. This is the quantitative form of the DECISIONS finding: a STATIC intraday stop cannot hold a swing."
    },
    {
      "id": "swing-migrating-stop-tournament-2026-08-06",
      "phase": "build",
      "status": "pending",
      "title": "Re-run the hold-rule tournament with MIGRATING stops (static stops cannot answer the swing question)",
      "owner": "claude",
      "priority": "P0",
      "property": "playbooks",
      "created_at": "2026-08-06",
      "path": "/z/trading/dev/swing-playbook-engine/census.py",
      "note": "86% of the 344 signals stop out inside the 10-session horizon, and the ATR sweep shows every static level losing to no stop. The charter already says 'STOP = swing invalidation, migrating with the trade'. Until the grid models a stop that moves with the trade, today's top-15 ranking is measuring the stop, not the hold rule."
    },
    {
      "id": "swing-sigma13-grid-member-2026-08-06",
      "phase": "fix",
      "status": "pending",
      "title": "Fix or replace the 1.3-sigma stop grid member - it defines only 47 of 344 signals and owns every survivor",
      "owner": "claude",
      "priority": "P0",
      "property": "playbooks",
      "created_at": "2026-08-06",
      "path": "/z/trading/dev/swing-playbook-engine/out/tournament.log",
      "note": "In census v1 the 1.3-sigma stop passed 0 of 21 (risk median 4.75 vs the engine's own 2 x ATR cap of 2.54). In v2 it is defined on only 47 of 344 signals - yet the entire BOTH-HALVES SURVIVORS list is stop_sigma_1_3 at n=47. That is a selection artifact: a stop that only exists on the 14% of signals with the tightest structure will look like an edge. Either replace the grid member or revisit maxStopATR, then re-score. Only hold_low and pine_literal are defined everywhere."
    },
    {
      "id": "swing-census-publish-page-2026-08-06",
      "phase": "ship",
      "status": "pending",
      "title": "Publish the swing census + tournament to a readable surface",
      "owner": "claude",
      "priority": "P1",
      "property": "playbooks",
      "created_at": "2026-08-06",
      "path": "/z/public_html/playbooks/",
      "note": "census_summary.json, funnel_by_symbol.csv, signals.csv, hold_rule_results.csv, stop_grid_chop_leak.csv and stop_sweep.csv all live inside /z/trading/dev/swing-playbook-engine/out/. Nothing is on a page. The funnel alone (11,969 -> 41,904 -> 949 -> 344) is worth a glance-able chart."
    },
    {
      "id": "hud-calendar-summary-modal-2026-08-06",
      "phase": "build",
      "status": "completed",
      "completed_at": "2026-08-06",
      "title": "HUD Calendar Summary modal (daily / weekly / monthly) + trader.php + render.js edits",
      "owner": "james",
      "priority": "P2",
      "property": "accountability",
      "path": "/z/trading/a1p_journal/accountability-v2/hud/js/calendar-modal.js",
      "note": "Reported by the nightly scan, not edited by it - the v2 cockpit stays read-only to this run. Real source edits landed in BOTH the dev and prod accountability-v2 trees today: render.js 10:44, api/trader.php 10:59, calendar-modal.js 17:05, hud/index.html 17:05, each with a timestamped .bak. calendar-modal.js provides a Calendar Summary modal with daily month-grid / weekly Mon-Fri / monthly modes, a custom Apply-range override, an in-modal trader dropdown, and UTC-noon date math to avoid DST drift, sourced from LA.api.calendar()."
    },
    {
      "id": "hud-prev-close-api-2026-08-07",
      "phase": "build",
      "status": "completed",
      "owner": "james",
      "priority": "P1",
      "property": "accountability",
      "project": "acct_v2",
      "title": "HUD prior-day-close API (api/prev-close.php) + render.js wiring, dev and prod",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/prev-close.php",
      "created_at": "2026-08-07",
      "completed_at": "2026-08-07",
      "note": "New batch prior-close endpoint backed by Yahoo v8 chart, same source chart-bars.php uses. GET ?symbols=A,B,C (max 40, regex-validated) -> {ok, closes{}, as_of}. Per-symbol file cache at api/cache/prev_close with a 300s TTL so the polling HUD does not hammer Yahoo. Carries a real fix in its own comments: Yahoo meta.previousClose / regularMarketPreviousClose is often NULL for ETFs and thin names, and meta.chartPreviousClose is the bar BEFORE the requested window (multi-day off) - so it walks the bars array backwards and returns the last close whose ET calendar day is not today, with a pre-market fallback to the newest close. Consumed in render.js at two render sites behind an in-flight guard and a client-side cache. Shipped to BOTH dev (17:53) and prod (17:55) with timestamped .bak files. render.js went through 7 saved revisions 12:52-19:17; main.js 10:43; hud/index.html 19:18 both trees. Reported only - cockpit stays read-only to the nightly run."
    },
    {
      "id": "hud-error-log-unreadable-2026-08-05",
      "phase": "build",
      "status": "pending",
      "owner": "james",
      "priority": "P1",
      "property": "accountability",
      "project": "acct_v2",
      "title": "accountability-v2 api/error_log unreadable over WebDAV - PHP errors written nightly, invisible",
      "path": "/z/trading/dev/a1p_journal/accountability-v2/api/error_log",
      "created_at": "2026-08-05",
      "note": "Sixth consecutive night. On 2026-08-09 the recursive scan confirmed FOUR PHP error sinks written on the day and none openable over WebDAV: accountability-v2/api/error_log on dev (23:10:29) AND prod (23:25:18), src/php/trade_reports/error_log (15:55:52), and src/php/tv_alerts_analytics/backend/storage/logs/php_error.log (23:11:33). Same ownership pattern throughout - the PHP/web user writes them, the WebDAV mount user cannot read them. Stakes rose today: the HTF Portfolio forks now depend on tradingview_smb_indicator_trigger.php accepting scorecard_tag, and if that receiver throws, this unreadable log is where it says so. Needs one local read over ssh a2."
    },
    {
      "id": "mgp-deliverability-probe-run-2026-08-08",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-08",
      "completed_at": "2026-08-08",
      "title": "MGP deliverability probe run - DMARC quarantine test + 3 mail-tester sends + direct-from-a2 send",
      "owner": "james",
      "priority": "P1",
      "property": "pm_system",
      "project": "atlas",
      "path": "/z/setup_daily_reporting/reports/notify.log",
      "note": "Five outbound probes logged in notify.log on 2026-08-08, all ok=true: 11:08 'DMARC quarantine test - check Show Original for auth results' to Jim.Domino@gmail.com (1272 B); 11:26 full Morning Game Plan render to test-afxk502tz@srv1.mail-tester.com (3917 B, em-dash subject with the lean-risk-on / breadth +2 / yellow lights payload); 11:40 same brief stripped to a plain-ASCII subject to test-bsqiwo2mp@ (1399 B); 13:55 third mail-tester probe to test-nuadjd4r8@ (1380 B); 14:05 'Test from a2 - direct server send' to Jim.Domino@gmail.com (446 B) isolating the a2 postfix path. The 3917 -> 1399 byte drop between the first two probes is the deliberate variable: MIME-encoded vs plain subject, the exact block named in the mgp-email-deliverability memory. Probes are done; the verdict is not recorded anywhere."
    },
    {
      "id": "mgp-deliverability-record-result-2026-08-08",
      "phase": "ship",
      "status": "pending",
      "created_at": "2026-08-08",
      "title": "Record the mail-tester score + DKIM/DMARC verdict from the 8/8 probes",
      "owner": "james",
      "priority": "P1",
      "property": "pm_system",
      "project": "atlas",
      "note": "STILL OPEN as written - no mail-tester score for afxk502tz / bsqiwo2mp / nuadjd4r8 has landed on disk, and the MIME-vs-plain-subject question is still unanswered. But the config change this task was meant to feed was decided and shipped anyway on 2026-08-09: see notify-smtp-transport-migration-2026-08-09, which moved both notifiers to authenticated SMTP from noreply@a1prospect.com (the sender the mgp-email-deliverability memory already records as scoring 10/10). What remains is the record-keeping, plus the still-unfinished cPanel DKIM Repair step from ams-email-deliverability-dmarc-diag. Lower stakes now than it was 24 hours ago."
    },
    {
      "id": "htf-portfolio-forks-shipped-2026-08-09",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-09",
      "completed_at": "2026-08-09",
      "title": "HTF Portfolio playbook shipped - Boat + VCET Pine forks, strategies.json registration, scorecard wiring, lens page",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "project": "classifier",
      "path": "/scorecards/htf-portfolio.html",
      "note": "New portfolio-hold playbook wired end to end on 2026-08-09, five layers. (1) Two Pine FORKS, parents left intact: lean_j3p/boat_portfolio_v1.0.pine.txt 18:32 (fork of THE BOAT v3.6f, adds Hold mode Day-scalp/Swing-HTF/Portfolio, true daily 5/10/20/50/200 MA bias + structure-stop ratchet, no LTF full-flat while HTF bias agrees, eventTag SMB_BOAT_PORT_EVT) and trading/indicators/src/indicators/Adaptive_Mean_Portfolio_V1.0.txt 20:07 (VCET fork). (2) strategies.json 15:36 registers htf-portfolio: family=level_breakout, featured+focus, setup_type=htf_portfolio, source_type=indicator_fork, bucket=portfolio, stats n=0, plus a matching tag entry in the classifier vocab. (3) gen_scorecard.py 18:29 adds HTF_PORTFOLIO to CANONICAL_SLUG + display map, a webhook slug passthrough, an entry_type fallback that maps BOAT/COMP/V-BOTTOM/NECK etc to HTF_PORTFOLIO, and a 5-row htf-portfolio checklist. (4) The WEBHOOK RECEIVER src/php/webhooks/tradingview_smb_indicator_trigger.php 18:29 (.bak-htf-strategy kept) now resolves $strategy = payload['strategy'] ?? payload['scorecard_tag'] ?? payload['playbook_type'] - the link that gets a fork's alert into indicator_alerts under the right tag, and the piece the first three scans missed. (5) /scorecards/htf-portfolio.html is the lens page; scorecards/index.html, system/index.html, STRATEGY_SCOREBOARD.html and _shared/jd_nav.js updated at 15:38. Forward basket: CRWD ORCL AVGO CRWV AMZN SNDK AMD NVDA MSFT TSLA. On-chart live A-F grade: A = open >= +1R and HTF-aligned, F = open against HTF bias."
    },
    {
      "id": "htf-portfolio-tv-compile-verify-2026-08-09",
      "phase": "validate",
      "status": "pending",
      "created_at": "2026-08-09",
      "title": "Compile-verify both Portfolio forks in TradingView and land one webhook under scorecard_tag=htf-portfolio",
      "owner": "james",
      "priority": "P1",
      "property": "strategy_library",
      "project": "classifier",
      "path": "/z/public_html/lean_j3p/boat_portfolio_v1.0.pine.txt, /z/trading/indicators/src/indicators/Adaptive_Mean_Portfolio_V1.0.txt",
      "note": "Both forks were written 2026-08-09 and neither has been loaded in TradingView. UPDATE 2026-08-16: Adaptive_Mean_Portfolio_V1.0.txt moved again at 20:21 ET - the FIFTH consecutive day the file has been edited without a compile check. Tonight the claim was tested directly rather than inferred: grepping the entire webhook_debug.log for Adaptive_Mean_Portfolio / htf-portfolio / HTF_PORT returns ZERO matches all-time. Not 'none today' - none ever. A webhook did arrive at 20:00 ET, 21 minutes before the edit, but it was a bare-text 'AVWAP WTD crossover on BTCUSD' alert that failed JSON decode, unrelated to the fork. So htf-portfolio is still registered at n=0, gen_scorecard.py's HTF_PORTFOLIO branch is still untested code, and the Test Before Ship rule has now been inverted five days in a row. Two things still to prove: (a) both scripts COMPILE clean on a portfolio layout, (b) one real alert reaches indicator_alerts with scorecard_tag=htf-portfolio."
    },
    {
      "id": "notify-smtp-transport-migration-2026-08-09",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-09",
      "completed_at": "2026-08-09",
      "title": "Email transport migrated to authenticated SMTP on both notifiers - mailer.php demoted to fallback",
      "owner": "Claude",
      "priority": "P1",
      "property": "pm_system",
      "project": "atlas",
      "path": "/z/setup_daily_reporting/reports/notify.py, notify_ams.py",
      "note": "Follow-on from the 8/8 deliverability probes. notify.py 13:10 and notify_ams.py 13:40 now try authenticated SMTP first (mail.a1prospect.com, sender noreply@a1prospect.com) and only fall through to the mailer.php webhook on failure, so a send is never dropped; .bak-presmtp copies of both kept alongside. mgp_secrets.json 13:10 carries the SMTP credentials. Four verified sends on the day: 12:44 MGP path test to 2 recipients, 13:11 'Pipeline test - SMTP transport' and 13:17 'New sender live - noreply@a1prospect.com' both logged 'via smtp mail.a1prospect.com' in notify.log, and 13:41 the AMS notifier's own test in notify_ams.log - the first SMTP line that log has ever carried. This is the config change the email-send-method and mgp-email-deliverability memories both specify, now actually in force in code rather than documented only."
    },
    {
      "id": "lynch-master-refresh-2026-08-09",
      "phase": "validate",
      "status": "completed",
      "created_at": "2026-08-09",
      "completed_at": "2026-08-09",
      "title": "LYNCH Master combinatorial backtest re-run - winner unchanged across two independent sweeps",
      "owner": "Claude",
      "priority": "P2",
      "property": "strategy_library",
      "path": "/system/LYNCH_MASTER.html",
      "note": "Rebuilt 22:00. This is a RE-RUN of the 8/02 sweep, not a new result: 3,612 combinations this time vs 3,618 on 8/02, 999 surviving the both-halves walk-forward test vs 998. The winner is byte-identical on both metrics - A+B+C / UVXY-mom x Reactive / Swing pure hold-5d / no gate, 435 trades, +$71,454, $14,992 max DD, 477% return-per-drawdown, 45.1% win, Sharpe 0.74. That stability across two independent sweeps is the actual news. Top-50 concordance also holds: grade 100%, sizing 84% (UVXY-mom x Reactive), while exit tops out at 26% and the gate splits 36/36/28 - so grade and sizing are the edge and exit/gate remain close to noise."
    },
    {
      "id": "gapma-scanner-build-2026-08-10",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-10",
      "completed_at": "2026-08-10",
      "title": "Build GAPMA - Gap Above 5MA scanner (longs + shorts) with full universe data layer",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "project": "goldline",
      "path": "/z/gt_dev/gapma/",
      "note": "New property built end to end 08:07-09:47. 1,284 per-symbol JSONs fetched 08:10-09:34, then index.json, earnings.json, longs.json (915 rows), shorts.json (370 rows). Each row carries gap_pct, open vs sma5, sma20 retrace 'better_trade' flag, MA slope + degrees for 5/10/20/50/200, RRG ETF + quadrant + points, industry, mktval, vol vs 30d. UI: index.html (longs), shorts.html, chart_pop.html, gapma.js, gapma-charts.js, gapma.css."
    },
    {
      "id": "pvp-ui-guide-build-spec-2026-08-10",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-10",
      "completed_at": "2026-08-10",
      "title": "Write the PVP UI guide + build spec (page regions, metrics, overlays, porting path)",
      "owner": "Claude",
      "priority": "P2",
      "property": "strategy_library",
      "project": "pvp",
      "path": "/z/gt_dev/pvp/guide.html",
      "note": "Documents page regions, QUALIFIED bottom-panel metrics, sidebar row fields, toolbar controls, chart overlays, alerts, full metric reference, layout, a 'Porting to Pullback Strategy' section and a build roadmap. index.html and pvp-charts.js republished alongside it at 22:03."
    },
    {
      "id": "vwap-band-research-page-2026-08-10",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-10",
      "completed_at": "2026-08-10",
      "title": "Publish the VWAP Band Research surface (core config baseline + alignment lens)",
      "owner": "Claude",
      "priority": "P1",
      "property": "param_surface",
      "project": "vwap-bands",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "note": "Generated 16:46. Core config = entry VWAP to -0.5SD, stop LB1, target UB3 (+2.0SD), 38 symbols over 10 trading days 3/18-4/2. n=390 core (3,510 all): 15.64% win, expectancy -0.2194R, PF 0.74, total -85.56R, avg win 3.94R vs avg loss 0.99R. env_tagged=0 (0.0%), so aligned_bull and not_aligned are both n=0 - the alignment split cannot be evaluated yet."
    },
    {
      "id": "team-chat-scrum-room-2026-08-10",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-10",
      "completed_at": "2026-08-10",
      "title": "Build the Team Chat scrum room (AI backend, API, voice input, config template)",
      "owner": "Claude",
      "priority": "P2",
      "property": "foundation",
      "project": "atlas",
      "path": "/z/gt_dev/chat/",
      "note": "Worked 07:45-10:15: .htaccess, ai.php, api.php, index.php, lib.php, voice.js, config.local.php.example. Titled 'Team Chat - Scrum Room', landing headline 'Scrum room for the day's work'."
    },
    {
      "id": "breadth-daemon-silent-launchers-2026-08-10",
      "phase": "ship",
      "status": "completed",
      "created_at": "2026-08-10",
      "completed_at": "2026-08-10",
      "title": "Add silent host-aware launchers + scheduler installer for the breadth daemon",
      "owner": "Claude",
      "priority": "P2",
      "property": "foundation",
      "project": "atlas",
      "path": "/z/gt_dev/breadth/",
      "note": "Written 09:08: install_home_scheduler.ps1, start_breadth_daemon.bat, start_breadth_daemon_silent.vbs, start_breadth_daemon_debug.bat, stop_breadth_daemon.bat, refresh_auto.bat. Matches the console-window-hiding convention (silent VBS wrapper, no flashing windows). Daemon wrote extended_live.json at 20:00."
    },
    {
      "id": "churn-fill-cap-per-name-2026-08-10",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-10",
      "title": "Cap fills per name - commission was the entire loss on 8/10",
      "owner": "james",
      "priority": "P0",
      "property": "scorecards",
      "project": "accountability",
      "path": "/z/public_html/scorecards/churn_today.json",
      "note": "8/10: 12 tickers, 704 fills, 8,143 shares. Gross -$15.28, commission $59.26, net -$74.53. Median 75.5 fills per name. TSLA 130 fills / 935 shares for -$43.69 gross. VZ made +$47.47 gross across 119 fills but paid $17.60 to keep +$29.87. A flat book turned into a loss purely on fill count."
    },
    {
      "id": "gapma-ranking-gate-2026-08-10",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-10",
      "title": "Add a ranking / permission gate to GAPMA before it earns a slot next to SIP",
      "owner": "Claude",
      "priority": "P1",
      "property": "strategy_library",
      "project": "goldline",
      "path": "/z/gt_dev/gapma/data/longs.json",
      "note": "At min_gap_pct 0.25 the scan returns 915 longs and 370 shorts - that is the tape, not a watchlist. Rank on fields already present: better_trade (low touched SMA20 within 10 sessions after being above it), RRG quadrant/points, MA slope degrees across 5/10/20/50/200, avdol_k and vol_vs_30 liquidity floors."
    },
    {
      "id": "mgp-v2-duplicate-send-guard-2026-08-10",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-10",
      "title": "MGP v2 sent 7 emails for 3 editions - the send guard and the autoheal detector disagree",
      "owner": "Claude",
      "priority": "P1",
      "property": "mgp",
      "project": "mgp",
      "path": "/z/setup_daily_reporting/reports/pipeline_autoheal.log",
      "note": "8/10 v2 sends: morning 06:47, 08:35 and 12:10; midday 11:39 and 11:43; recap 16:41 and 16:48. Autoheal fired mgp_v2_morning at 12:06 because /mgp/v2/2026-08-10.html was missing - so the sender thinks it ran while the healer keys on a file that was never written until the heal. A write failure produces duplicate emails instead of a silent repair. v1 editions were clean: one send each at 05:42, 11:30, 16:40."
    },
    {
      "id": "james-ver-vcet-portfolio-fork-2026-08-10",
      "phase": "design",
      "status": "in_progress",
      "created_at": "2026-08-10",
      "title": "James's own cut of the VCET Portfolio fork (James-Ver-Adaptive_Mean_Portfolio_V1.0)",
      "owner": "james",
      "priority": "P1",
      "property": "strategy_library",
      "project": "classifier",
      "path": "/z/trading/indicators/src/indicators/James-Ver-Adaptive_Mean_Portfolio_V1.0.txt",
      "note": "Written 19:46 alongside yesterday's Adaptive_Mean_Portfolio_V1.0.txt. Not yet compile-verified in TradingView and no webhook has landed under htf-portfolio (still n=0). The BOAT-side fork in lean_j3p was not touched today."
    },
    {
      "id": "ams-ib-stops-mvp-spec-2026-08-11",
      "phase": "design",
      "status": "completed",
      "created_at": "2026-08-11",
      "completed_at": "2026-08-11",
      "title": "Write the IB stops MVP spec for AMS",
      "owner": "james",
      "priority": "P1",
      "property": "ams",
      "project": "ams",
      "path": "/z/trading/dev/ams/docs/IB-STOPS-MVP.md",
      "note": "New spec doc landed 08-11 10:50. First written artifact of the AMS broker-connectivity lane; the brokers.php + _api endpoints that followed on 08-13/08-14 build against it."
    },
    {
      "id": "vcet-portfolio-fork-port-v11-2026-08-11",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-11",
      "completed_at": "2026-08-11",
      "title": "Cut VCET Portfolio Unified port_v1.1 and the James-Ver portfolio fork",
      "owner": "james",
      "priority": "P1",
      "property": "strategies",
      "project": "strategy-library",
      "path": "/z/gt_dev/jd/VCET_Portfolio_Unified_port_v1.1.txt",
      "note": "Two files 08-11 evening: James-Ver-Adaptive_Mean_Portfolio_V1.0.txt at 19:39, VCET_Portfolio_Unified_port_v1.1.txt at 20:55. Merges the HTF/Portfolio hold with the James-Ver fast-EMA exits. Written, not compile-verified."
    },
    {
      "id": "am-vptags-helper-2026-08-12",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-12",
      "completed_at": "2026-08-12",
      "title": "Add the _am_vptags volume-profile tag helper",
      "owner": "james",
      "priority": "P2",
      "property": "strategies",
      "project": "strategy-library",
      "path": "/z/gt_dev/jd/_am_vptags.py",
      "note": "Landed 08-12 19:07 alongside the Adaptive Mean portfolio work."
    },
    {
      "id": "ams-brokers-page-2026-08-13",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-13",
      "completed_at": "2026-08-13",
      "title": "Build the AMS brokers page",
      "owner": "james",
      "priority": "P1",
      "property": "ams",
      "project": "ams",
      "path": "/z/trading/dev/ams/app/brokers.php",
      "note": "08-13 14:44. The surface the 08-14 broker API batch hangs off."
    },
    {
      "id": "ams-broker-api-layer-2026-08-14",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-14",
      "completed_at": "2026-08-14",
      "title": "Ship the AMS broker API layer - orders, positions, quotes, chains, order intent",
      "owner": "james",
      "priority": "P0",
      "property": "ams",
      "project": "ams",
      "path": "/z/trading/dev/ams/_api/",
      "note": "Eight files in one 08:08 batch on 08-14: _api/broker_orders.php, broker_positions.php, equity_quote.php, option_chain.php, order_intent.php plus app/positions.php, positions_data.php, symbol_executions.php. Biggest single AMS push of the week. Endpoints exist; they are also writing to error_log the same day - see ams-broker-api-errors-2026-08-14."
    },
    {
      "id": "vcet-portfolio-unified-port-v113-2026-08-14",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-14",
      "completed_at": "2026-08-14",
      "title": "Iterate the Portfolio fork to port_v1.13",
      "owner": "james",
      "priority": "P1",
      "property": "strategies",
      "project": "strategy-library",
      "path": "/z/gt_dev/jd/Adaptive_Mean_Portfolio_V1.0.txt",
      "note": "08-14 08:59. Now VER port_v1.13 - three display modes (Full / Clean / Stealth), regime background tint, MA state cloud on the 50-min rail, rail-slope lookback. Engine identical across modes, only drawing changes. Still not compile-verified in TradingView."
    },
    {
      "id": "boat-v14-neckline-rework-2026-08-14",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-14",
      "completed_at": "2026-08-14",
      "title": "Rebuild THE BOAT trigger around the neckline instead of the pre-decline swing high",
      "owner": "james",
      "priority": "P1",
      "property": "strategies",
      "project": "strategy-library",
      "path": "/z/gt_dev/jd/BOAT_v36g.pine.txt",
      "note": "08-14 09:43, BOAT v1.4. Every earlier version used the swing high BEFORE the decline as the deck - a level that exists before the boat does. v1.4 builds the deck from the RISE, which moves entry much closer to the keel; the old left-rim trigger sat so far above that it kept tripping the CHASING gate at 2+ ATR from the EMA."
    },
    {
      "id": "vwap-band-research-index-2026-08-14",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-14",
      "completed_at": "2026-08-14",
      "title": "Publish the VWAP Band Research System index",
      "owner": "james",
      "priority": "P2",
      "property": "backtests",
      "project": "vwap-band-entry-surface",
      "path": "/z/gt_dev/db_reports/vwap-research/index.html",
      "note": "08-14 16:46. Front door for the VWAP band entry-surface research lane."
    },
    {
      "id": "ams-broker-api-errors-2026-08-14",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-14",
      "title": "Fix the PHP errors the new AMS broker endpoints are throwing",
      "owner": "james",
      "priority": "P0",
      "property": "ams",
      "project": "ams",
      "path": "/z/trading/dev/ams/_api/error_log",
      "note": "The five broker endpoints landed 08-14 08:08. app/error_log written 16:26, _api/error_log written 19:23 the same day. Read both logs before extending the layer."
    },
    {
      "id": "alert-subject-sanitize-2026-08-13",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-13",
      "title": "Sanitize alert subject lines before they reach email and SMS",
      "owner": "james",
      "priority": "P1",
      "property": "system",
      "project": "sip-alert-pipeline",
      "path": "/z/setup_daily_reporting/reports/notify.log",
      "note": "notify.log line 199, 08-13 08:41:58: a test subject containing shell metacharacters was delivered verbatim to jdomino@smbcap.com and 7326886208@vtext.com. Reads as your own probe of the alert path, and it answered the question - the subject is not escaped. Real headlines carrying quotes, semicolons or backticks ride the same road. Nothing was executed; the string was only ever mail-header text."
    },
    {
      "id": "mgp-v1-morning-miss-2026-08-13",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-13",
      "title": "Find out why the 08-13 morning MGP v1 never emailed and autoheal did not recover it",
      "owner": "james",
      "priority": "P1",
      "property": "mgp",
      "project": "mgp-pipeline",
      "path": "/z/setup_daily_reporting/reports/state/autoheal_mgp_v1_morning_2026-08-13.tries",
      "note": "notify.log has no 'Domino Brief - Morning Game Plan 2026-08-13' entry - 08-11, 08-12 and 08-14 all have one. MGP v2 sent normally at 06:47, midday 11:30 and recap 16:30 both fine, so only the v1 morning edition was lost. The autoheal tries file was written at 12:05, five hours after the edition was due and long after it mattered."
    },
    {
      "id": "nightly-pm-three-night-gap-2026-08-13",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-14",
      "title": "Stop the nightly PM from silently skipping nights when the machine sleeps",
      "owner": "james",
      "priority": "P1",
      "property": "todos",
      "project": "nightly-pm",
      "path": "/z/logs/nightly-startup.log",
      "note": "last_nightly_run sat at 2026-08-10T23:06 until tonight - three consecutive nights (08-11, 08-12, 08-13) never fired. Backfilled by the 08-14 run. The cron is correct; the machine is asleep at 23:00. Nothing surfaces this except a human noticing the calendar has holes."
    },
    {
      "id": "nightly-scan-webdav-stall-2026-08-14",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-14",
      "title": "Route the accountability-v2 nightly scan through ssh a2 instead of WebDAV",
      "owner": "james",
      "priority": "P1",
      "property": "todos",
      "project": "nightly-pm",
      "path": "/z/setup_daily_reporting/reports/scan_recent_activity.py",
      "note": "Scan A2 took 28 min on 08-14 and ~35 min on 08-15. UPDATE 2026-08-16: tonight it did not return at all - all three scan_recent_activity.py calls plus the heartbeat were still hung after ~20 minutes and were killed. The nightly ran to completion only because the whole scan was re-routed server-side. That fallback is now proven, not theoretical: a single `ssh a2 find ... -newermt` over the same 19 roots returned the full day's file list in a few SECONDS, against 20+ minutes of stall over WebDAV. Raising to P1 - this is no longer a slow-path annoyance, it is the difference between the nightly having evidence and not. Fold the ssh walk into scan_recent_activity.py as the default with the WebDAV walk as fallback, per feedback_ssh_for_zdrive_scans. Reference implementation used tonight is in the session scratchpad (ssh_scan.py): one find over ~/setup_daily_reporting, ~/gt_dev/db_reports, ~/trading/..., ~/public_html/* with -printf '%TY-%Tm-%Td %TH:%TM:%TS  %p'."
    },
    {
      "id": "hud-api-error-logs-2026-08-14",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-14",
      "title": "Read the HUD error logs - prod accountability-v2 API is writing errors",
      "owner": "james",
      "priority": "P3",
      "property": "accountability",
      "project": "accountability-v2",
      "path": "/z/trading/a1p_journal/accountability-v2/api/error_log",
      "note": "CORRECTED 2026-08-16. The 08-14 and 08-15 writeups both reported this as 'the production HUD is writing errors'. That was wrong, and the correction matters because it was pointing at a bug that is not there. The actual tail of both prod and dev accountability-v2/api/error_log is: '[17-Aug-2026 02:27:05 UTC] DB Config / Database : trsports_tradingjournal_dev / DB User : trsports_trader / Host : localhost' followed by 'Database connection established successfully.' That is informational connect chatter being written to error_log, i.e. a log-LEVEL problem, not a runtime failure - and it is already tracked as hud-error-log-db-config-spam-2026-08-02. Dropping P1 to P3 and folding into that item. Two real details survive: the log stamps UTC with a UTC suffix while everything else is ET, and prod is pointed at a database named trsports_tradingjournal_DEV, which is worth one look. HUD stays read-only per feedback_hud_off_limits - reported, not touched."
    },
    {
      "id": "j3p-simple-version-2026-08-15",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P1",
      "property": "system",
      "project": "exec",
      "title": "The Simple Version - plain-language explainer of the Path to Green finding",
      "path": "/z/public_html/system/J3P_SIMPLE_VERSION.html",
      "created_at": "2026-08-15",
      "completed_at": "2026-08-15",
      "note": "New page at 17:53. States the whole diagnosis in words a non-trader can follow: 205 trades/day at ~5 shares, 12,306 trades over 60 days, 43% win, +$3.75 avg win vs -$3.20 avg loss = -21 cents per trade. Bar chart of expectancy by trade number (2nd-3rd +$2.30, 6th-10th +$0.97, then negative through 151st+), worst block is trades 41-80 at -$1,087. Four ordered steps (cap at 10 -> then size up -> then hold winners -> then the Management Doctrine) with an explicit warning about why sizing up FIRST loses faster. Also lists the three refuted ideas. Links to J3P_MANAGEMENT_DOCTRINE.html."
    },
    {
      "id": "path-to-green-narrative-refresh-2026-08-15",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "property": "system",
      "project": "exec",
      "title": "Path to Green + Management Doctrine pages refreshed alongside the Simple Version",
      "path": "/z/public_html/system/J3P_PATH_TO_GREEN.html",
      "created_at": "2026-08-15",
      "completed_at": "2026-08-15",
      "note": "J3P_MANAGEMENT_DOCTRINE.html rewritten 16:16, J3P_PATH_TO_GREEN.html 17:27, then J3P_SIMPLE_VERSION.html 17:53 - one continuous Saturday pass over the three narrative surfaces, ending in the plain-language cut."
    },
    {
      "id": "classify-trades-reclass-backfill-2026-08-15",
      "phase": "build",
      "status": "completed",
      "owner": "claude",
      "priority": "P2",
      "property": "scorecards",
      "project": "classifier",
      "title": "classify_trades.py edited and a reclassification backfill run over 07-17..08-11",
      "path": "/z/setup_daily_reporting/reports/classify_trades.py",
      "created_at": "2026-08-15",
      "completed_at": "2026-08-15",
      "note": "Source edited 14:48:56; the run followed 14:55-14:57 and rewrote roughly 420 files under reports/cache/classify_trades_bars/ covering sessions 2026-07-17 through 2026-08-11. That is a --refresh reclassification, not an incremental fill. Worth confirming the slug distribution actually moved - the whole point of a refresh is that the new classifier disagrees with the old rows, and nothing in tonight's scan proves it did."
    },
    {
      "id": "eod-multitrader-crash-2026-08-15",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "property": "scorecards",
      "project": "multitrader",
      "title": "EOD catch-up multi-trader step crashes with a traceback",
      "path": "/z/setup_daily_reporting/reports/eod_catchup.log",
      "created_at": "2026-08-15",
      "note": "2026-08-15T16:45:03 'mt ERROR: Traceback (most recent call last):' - only the first line reached the log, so the exception was lost. UPDATE 2026-08-16: the failure changed shape and got HARDER to diagnose, not easier. Tonight's 16:45 run logged no 'mt ERROR' line at all AND no closing 'catch-up done' line - the log simply stops after 'WROTE: you-vs-you.html' at 16:45:14. But the step did partially run: it wrote scorecards/traders/traders.json at 16:45:18 and exactly one card, traders/james/trade_scorecard_2026-08-12.html, at 16:47:06, then died with nothing on stdout. So it now gets through the roster (all 6 traders present in traders.json) and one historical card before dying silently. Fix the logging FIRST - two nights running, the evidence needed to fix the bug is the thing that is missing. Same run repeats the known identity failure (relation alerts_tradingview.users_accounts does not exist, defaults to user_id=2)."
    },
    {
      "id": "trade-count-cap-live-meter-2026-08-15",
      "phase": "design",
      "status": "pending",
      "owner": "claude",
      "priority": "P0",
      "property": "system",
      "project": "exec",
      "title": "Live trade-count meter enforcing the 10-trade daily cap",
      "path": "/z/public_html/system/J3P_LIVE.html",
      "created_at": "2026-08-15",
      "note": "The Simple Version's step 1 is a hard stop at 10 trades/day, backtested at -$2,104 -> +$504 with fees falling ~$2,000 -> ~$110. It currently lives only as prose. Needs to be a live count-down that auto-populates from fills mid-session, loud when it is the changing panel, and it must NOT be built into the live-accountability-v2 cockpit (read-only). Build it on a J3P surface. Step 2 (sizing up) stays blocked until this is running - sizing up while expectancy is negative loses faster."
    },
    {
      "id": "hud-trades-api-promotion-2026-08-15",
      "phase": "build",
      "status": "completed",
      "owner": "james",
      "priority": "P2",
      "property": "accountability_hud",
      "project": "acct-v2",
      "title": "HUD trades API + scorecard-click promoted dev to prod",
      "path": "/z/trading/a1p_journal/accountability-v2/api/trades.php",
      "created_at": "2026-08-15",
      "completed_at": "2026-08-15",
      "note": "api/trades.php dev 14:51 -> prod 14:56; hud/js/scorecard-click.js dev 14:53 -> prod 15:05; hud/index.html both trees 14:58. Dev-first then prod a few minutes behind - a deliberate promotion, not a stray edit. Landed three minutes after the classify_trades.py edit at 14:48, so classifier and cockpit moved as one push. Reported only: the cockpit is read-only from the agent side per the HUD rule."
    },
    {
      "id": "acct-v2-daemon-scripts-refresh-2026-08-15",
      "phase": "build",
      "status": "completed",
      "owner": "james",
      "priority": "P3",
      "property": "accountability_hud",
      "project": "acct-v2",
      "title": "Daemon launcher + watchdog + switch-machine scripts refreshed in both trees",
      "path": "/z/trading/a1p_journal/accountability-v2/scripts/watchdog.ps1",
      "created_at": "2026-08-15",
      "completed_at": "2026-08-15",
      "note": "silent-launcher.vbs, watchdog.ps1, switch-machine.ps1 - all six files across dev and prod between 17:46:40 and 17:46:46. Host-aware launcher and watchdog maintenance."
    },
    {
      "id": "hud-prod-error-log-2026-08-15",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P1",
      "property": "accountability_hud",
      "project": "acct-v2",
      "title": "Production HUD api/error_log still being written",
      "path": "/z/trading/a1p_journal/accountability-v2/api/error_log",
      "created_at": "2026-08-15",
      "note": "Touched 23:37 tonight during the nightly run - same signature as 08-14, and now on the same day api/trades.php was promoted to prod. Read the log before assuming the promotion is clean. Read-only per the HUD rule: report, do not edit."
    },
    {
      "id": "classify-refresh-verify-slugs-2026-08-15",
      "phase": "build",
      "status": "pending",
      "owner": "claude",
      "priority": "P2",
      "property": "scorecards",
      "project": "classifier",
      "title": "Verify the 07-17..08-11 reclassification actually changed slug assignments",
      "path": "/z/setup_daily_reporting/reports/classify_trades.py",
      "created_at": "2026-08-15",
      "note": "Today's --refresh rewrote ~420 bar-cache files. A refresh only earns its runtime if the new classifier disagrees with the old rows. Diff the trade_strategy_v2 slug distribution before and after; if it is unchanged, the edit at 14:48 did not do what it was meant to."
    },
    {
      "id": "lynch-master-refresh-2026-08-16",
      "phase": "validate",
      "status": "completed",
      "created_at": "2026-08-16",
      "completed_at": "2026-08-16",
      "title": "LYNCH Master combinatorial backtest re-run - third sweep, winner byte-identical again",
      "owner": "claude",
      "priority": "P2",
      "property": "strategy_library",
      "path": "/system/LYNCH_MASTER.html",
      "note": "Rebuilt 22:00 (31.8KB). Third sweep of the same grid: 3,612 combinations, 996 surviving the both-halves walk-forward test, vs 3,612/999 on 08-09 and 3,618/998 on 08-02. Winner is unchanged and identical to the dollar - A+B+C grade / UVXY-mom x Reactive sizing / Swing pure hold-5d exit / no gate, 435 trades, +$71,454, $14,992 max drawdown. Top-50 concordance also repeats exactly: grade 100% (50/50), sizing 84% (42/50 UVXY-mom x Reactive), exit tops out at 26% (hold-5d) and the gate splits 36/36/28. So the durable read across three sweeps is the same one: SIZING is the edge, exit and gate are noise. The winning recipe is highly selective - 435 trades over 5 years is ~87/yr, which is the selectivity thesis in a different asset class. See lynch-window-anchored-verify."
    },
    {
      "id": "lynch-window-anchored-verify-2026-08-16",
      "phase": "validate",
      "status": "pending",
      "created_at": "2026-08-16",
      "title": "Check whether the LYNCH backtest window is anchored - a week of new data moved nothing",
      "owner": "james",
      "priority": "P2",
      "property": "strategy_library",
      "path": "/system/LYNCH_MASTER.html",
      "note": "The 08-16 sweep returns the winner's trade count (435), total (+$71,454) and max drawdown ($14,992) byte-identical to the 08-09 sweep, seven calendar days and five trading days later. A rolling 5-year window should have added bars and moved at least the trade count. Two candidates: (a) the window end date is hard-coded, so re-running is a no-op, or (b) the sweep is serving a cached result. The one number that DID move is the survivor count, 999 -> 996, which is marginal recipes flipping around the both-halves threshold - that is consistent with a small data change but not with 435 trades staying frozen. Worth ten minutes before anyone runs a fourth sweep: confirm the end date, and if it is anchored, either unanchor it or stop re-running it and label the page with its as-of date."
    },
    {
      "id": "j3p-week-ahead-2026-08-16",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-16",
      "completed_at": "2026-08-16",
      "title": "J3P Week Ahead built for the 08-17 week - Friday review plus 24 names to prepare",
      "owner": "claude",
      "priority": "P1",
      "property": "system",
      "project": "exec",
      "path": "/z/public_html/system/J3P_WEEK_AHEAD.html",
      "note": "Written 10:00 Sunday, 537KB - by far the largest page in /system/. Two sections: '1 - How Friday actually went (2026-08-14)' and '2 - Names to prepare (24)'. This is the Sunday prep surface feeding Monday's open. Size is worth a look on its own: at half a megabyte it is ~17x the LYNCH page, which suggests per-name chart or bar data is being inlined rather than fetched, and that is a load-time cost on the one page he opens before the open."
    },
    {
      "id": "accountability-staging-entity-2026-08-16",
      "phase": "build",
      "status": "completed",
      "created_at": "2026-08-16",
      "completed_at": "2026-08-16",
      "title": "Isolated staging workspace stood up for the smb-reports-v2 accountability reports",
      "owner": "james",
      "priority": "P2",
      "property": "accountability",
      "project": "acct-v2",
      "path": "/z/gt_dev/db_reports/smb-reports-v2/accountability-staging-entity/",
      "note": "~30 files written 20:25:10-20:25:32 across accountability/ and a new sibling accountability-staging-entity/. This is the Parallel-Build Pattern applied to the reports tree: staging gets its own index-staging.html (sets window.__ACC_DATA_BASE__='staging' so the UI reads staging/data*.json only), its own update-staging.js writing under staging/, and promote-from-staging.ps1 / promote-html-to-prod.ps1 / promote-staging-to-prod.ps1 to ship it forward with backups under backups/promote-pre-* and a promotion-log.csv. STAGING-PROMOTE.md documents the roles. Net effect: the accountability reports now have a dev lane that cannot touch prod data, which is exactly what ADR-0001 asks for. The 22-second write burst says this was a scripted sync, not hand edits."
    },
    {
      "id": "webhook-log-utc-timestamps-2026-08-16",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-16",
      "title": "Webhook debug log stamps UTC while the rest of the system is ET",
      "owner": "james",
      "priority": "P2",
      "property": "alerts_hotkeys",
      "project": "classifier",
      "path": "/z/trading/dev/a1p_journal/src/php/webhooks/webhook_debug.log",
      "note": "The receiver writes '[2026-08-17 00:00:25]' for an event whose file mtime is 2026-08-16 20:00:25 ET - the log is stamped UTC, four hours ahead, with no suffix saying so. This breaks the Times-Are-Eastern rule and it is not cosmetic: it cost real time tonight, because grepping the log for today's date returned nothing and the first read of the evening concluded there had been no webhook traffic. Anything that joins this log to an ET-keyed table by date is silently off by one for the 20:00-24:00 ET window. Fix is one date() call in the logger plus an ' ET' suffix."
    },
    {
      "id": "webhook-plaintext-alert-rejects-2026-08-16",
      "phase": "design",
      "status": "pending",
      "created_at": "2026-08-16",
      "title": "Bare-text TradingView alerts are hitting the webhook and being rejected as invalid JSON",
      "owner": "james",
      "priority": "P3",
      "property": "alerts_hotkeys",
      "project": "classifier",
      "path": "/z/trading/dev/a1p_journal/src/php/webhooks/tradingview_smb_indicator_trigger.php",
      "note": "Of the last 400 payloads: 330 SMB_BOAT_EVT, 30 'NVDA Crossing 225.30', 12+10+8+2 SMB_FASH_EVT_Cursor_, 6 SMB_FASH_EVT_Office_Warriors, 1 'AVWAP WTD crossover on BTCUSD' (tonight, 20:00 ET), 1 '1-day AVWAP cross on TSLA'. So roughly 32 of 400 are plain-text alerts that fail JSON decode and return 'status=error msg=Invalid JSON payload'. They are alerts someone set up in TradingView without the TAG|{json} contract - they fire, they cost a webhook slot, and they land nowhere. Either convert them to the contract or delete them in TV; right now they are pure noise in the one log used to prove whether a strategy is emitting."
    },
    {
      "id": "nightly-heartbeat-webdav-hang-2026-08-16",
      "phase": "build",
      "status": "pending",
      "created_at": "2026-08-16",
      "title": "Nightly heartbeat hung on the WebDAV append, and the DONE line records a mangled report path",
      "owner": "james",
      "priority": "P2",
      "property": "todos",
      "project": "nightly-pm",
      "path": "/z/setup_daily_reporting/reports/heartbeat.py",
      "note": "Two separate defects in the tripwire itself. (1) Tonight's 'heartbeat.py LAUNCHED' never returned - it was still hung after ~20 minutes and had to be killed, so no LAUNCHED line was written and the silent-failure tripwire was blind on the night it was needed. (2) Every DONE line in /z/logs/nightly-startup.log records report=C:/Program Files/Git/todos/daily/<date>.html - Git Bash MSYS path conversion is rewriting the leading-slash '/todos/daily/...' argument into a Windows path before Python sees it. The log is also missing 08-11, 08-12 and 08-13 entirely even though daily reports exist for all three, so its coverage is already unreliable. Both fixes are small: write the heartbeat server-side over ssh instead of through WebDAV, and pass the report path in a form MSYS will not rewrite."
    }
  ],
  "projects": [
    {
      "id": "atlas",
      "label": "Atlas",
      "title": "Mission Control (PM platform)",
      "lead": "claude"
    },
    {
      "id": "ams",
      "label": "AMS",
      "title": "Auth + Access Control",
      "lead": "claude"
    },
    {
      "id": "exec",
      "label": "Exec",
      "title": "Execution Pipeline",
      "lead": "tbd"
    },
    {
      "id": "acct_v2",
      "label": "Acct-v2",
      "title": "Accountability HUD v2",
      "lead": "jimbo"
    },
    {
      "id": "classifier",
      "label": "Classifier",
      "title": "Live Classifier productionization",
      "lead": "roberto"
    },
    {
      "id": "multi_trader",
      "label": "MultiTrader",
      "title": "Multi-trader rollout (Phase 4)",
      "lead": "claude"
    },
    {
      "id": "strat_lib",
      "label": "StratLib",
      "title": "Strategy Library expansion",
      "lead": "alex"
    },
    {
      "id": "mtfvp",
      "label": "MTFVP",
      "title": "MTF VP Scanner Path B",
      "lead": "roberto"
    },
    {
      "id": "backup",
      "label": "Backup",
      "title": "JIMDO warm-backup runner",
      "lead": "claude"
    },
    {
      "id": "sip",
      "label": "SIP",
      "title": "Stocks-in-Play composite scanner",
      "lead": "claude",
      "description": "3-pillar daily watchlist: catalyst + RRG + daily movers. Gates Trigger Listener so alerts only fire on names with buyers present."
    },
    {
      "id": "pinepb",
      "label": "PinePB",
      "title": "Pine Playbooks Indicator (3-play TradingView indicator + webhook)",
      "lead": "claude",
      "description": "Single TradingView Pine v5 indicator implementing VWAP Pull-In, Trend Trade, Back$ide Scalp with toggleable plays. Fires alerts via webhook into alerts_tradingview.indicator_alerts (same shape as FashLate). Becomes the entry-trigger source for Sprint 2+ when the 3 named playbooks need their own indicators rather than borrowed from FL2 family."
    }
  ],
  "sprints": [
    {
      "id": 1,
      "label": "Sprint 1",
      "title": "Foundation hardening + AMS unblock",
      "start_date": "2026-06-22",
      "end_date": "2026-06-26",
      "status": "active",
      "goal": "Close Phase 0 gaps. Seed AMS users so onboarding stops being painful. Fix Acct-v2 date-range bugs."
    },
    {
      "id": 2,
      "label": "Sprint 2",
      "title": "Classifier production + AMS Phase 1 sign-off",
      "start_date": "2026-06-29",
      "end_date": "2026-07-03",
      "status": "queued",
      "goal": "Get the classifier refreshing live (unblocks Exec). Close AMS Phase 1."
    },
    {
      "id": 3,
      "label": "Sprint 3",
      "title": "Exec Trigger Engine + Alert Channels",
      "start_date": "2026-07-06",
      "end_date": "2026-07-10",
      "status": "queued",
      "goal": "First half of execution arm. Alerts to humans only \u2014 no orders yet."
    },
    {
      "id": 4,
      "label": "Sprint 4",
      "title": "Exec Gr8Trade adapter + forward-test",
      "start_date": "2026-07-13",
      "end_date": "2026-07-24",
      "status": "queued",
      "goal": "First broker adapter, forward-test only."
    },
    {
      "id": 5,
      "label": "Sprint 5",
      "title": "Live Gr8Trade execution + StratLib expansion",
      "start_date": "2026-07-27",
      "end_date": "2026-08-07",
      "status": "queued",
      "goal": "The milestone where the system starts paying for itself. Real Gr8Trade orders behind conservative caps."
    },
    {
      "id": 6,
      "label": "Sprint 6",
      "title": "Multi-broker + AMS Phase 5 promotion",
      "start_date": "2026-08-10",
      "end_date": "2026-08-28",
      "status": "queued",
      "goal": "Scale broker coverage and complete AMS migration."
    }
  ],
  "active_sprint_id": 1
}
