Morning Brief · ATLAS Standup

Saturday 2026-08-01 2:43 PM ET  ·  published 2:43 PM ET  ·  scheduled for 7:45 AM ET. All times America/New_York (ET).
3
shipped 7/31
29
stale blockers
39
open P0
74
open P1
7
in flight
4
blocked
3
dead monitors

What shipped overnight completed 2026-07-31 (Fri) ET

Provenance: these three tickets were written by the nightly PM at 2:12 PM ET today, not at 11 PM Friday — they are part of the same wake-burst that fired this brief. The nightly had not run since 7/28, so 7/29, 7/30 and 7/31 were all backfilled in one pass. The underlying files are real and timestamped Friday (SENTIMENT_ACCURACY.html 08:53, J3P_CUP_VERIFY 20:26, EOD backfill 16:45 — ok=18 fail=0); only the bookkeeping is late.

Decisions waiting on James status = blocked · nothing moves until these are called

Friday’s nightly PM flagged both of these as “untouched for a third day.” They are each a yes/no, not a build.

Stale blockers owner ≠ Claude · pending · P0/P1 · age ≥ 7 d

29 total
user 18
jimbo 4
alex 3
james 3
roberto 1
oldest 106 d
AgeProjectOwnerPriTitle
106 dAtlasuserP06/6 indicator_alerts DDL audit: dump schema, commit as ground truth
105 dMultiTraderuserP0USER: 3 steps to finish journal persistence
105 dMultiTraderuserP1Re-run gen_scorecard.py ON a2 (or via tunnel) and confirm enrichment populates
102 dAtlasuserP1Deploy z_activity_scanner.py on each Windows machine via Task Scheduler
101 dMultiTraderuserP0USER: run scientific_layer migrations 001–009 via psql tunnel
87 dAtlasuserP1Cron /system/update_freshness.py to keep freshness strip auto-refreshing
87 dMultiTraderuserP1Trader-of-the-Week scorecard review session (rotating: Nolan / Preston / Brian / Thomas)
87 dMultiTraderuserP1First-scorecard feedback session with each of the 4 traders
87 dStratLibalexP1Codify Range-Break strategy rules from Alex's Notion (entry / stop / target / best-when)
87 dStratLibalexP1Range-Break: walk-forward backtest published at /backtests/range-break/
87 dStratLibalexP1Register Range-Break in /strategies/ Strategy Library + canonical catalog
87 dClassifierrobertoP1Live Classifier HUD: auto-refresh classify/<SYMBOL>.json during market hours
87 dClassifieruserP1Review Roberto's Claude artifact + slot his real work into the right property
86 dAcct-v2jimboP1Move accountability-v2 Node services from nohup → Windows Task Scheduler (production-grade)
86 dAcct-v2jimboP1Wire v1 fills-poller to also POST into v2 /api/ingest.php (dual-write transition)
86 dAcct-v2userP1Upload AccountabilityTickBridge.py to KITE forward-test + verify ticks flow into marks_live
86 dAcct-v2userP1Trade Action Requests feature — finalize (close/stop intent flow: hud-orderlab + migration 021 + trade-action.php)
86 dAcct-v2userP1Execute dev→prod promotion sequence per docs/PROMOTE-DEV-TO-PROD.md
86 dAcct-v2jimboP0BUG: Risk Bump Progress bar does not respect the selected date range (always shows 5-day rolling, even when range = full month)
85 dAtlasuserP1Test mklink workaround for Claude Code 'suspicious path pattern' prompt on WebDAV-backed projects (unblocks nightly PM routine + agentic sessions)
85 dAcct-v2userP1Decide whether Risk Manager HUD (/accountability/risk.html, mock) is redundant with Accountability HUD (/live-accountability-v2/hud/) — merge or retire
80 dAtlasjamesP0Compliance follow-up: delete (or hide) historical MGP pages 4/17 through 5/12 — they still contain Vital Knowledge text
80 dAcct-v2jimboP0BUG: Ticket Meter denominator doesn't scale with date range (shows 1901 / 30 on 13-day MTD range)
78 dAtlasJamesP0BUG: MGP auto-fire skipped 5/14 mid-day + recap; 5/15 morning ran ~2h45m late
70 dAtlasjamesP0Nightly PM daemon went dark 5/20-5/22 (3 consecutive missed fires) — fix wake/launch
42 dAMSuserP0AMS: seed main.users with trader rows so identity_fn returns non-zero
42 dAMSuserP0AMS: resolve 7 open decisions D1-D7 (canonical domain · auth stack · DB strategy · sharing · risk_hud · 2FA · daily-stop)
41 dPinePBuserP0TradingView setup guide for James — paste Pine script, configure alerts, test
41 dPinePBuserP0James pastes finalized Pine script into TradingView + configures the 3 alerts

Open P0s by project 39 open P0 across 8 projects

Atlas 13
SIP 7
MultiTrader 4
AMS 4
PinePB 4
Untagged 3
Acct-v2 2
StratLib 1
Classifier 1
AgeProjectOwnerStatusTitle
106 dAtlasclaudepending1/6 Data-freshness dashboard: /system/freshness.json + banner on hub
106 dAtlasclaudepending2/6 Pipeline orchestrator: run_nightly.py calls every generator, logs to one place, fails loud
106 dMultiTraderclaudepending3/6 main.daily_summary SQL view: kills duplicate computation
106 dAMSclaudepending4/6 Identity function: resolve_user(account_id) -> (user_id, trader_name)
106 dStratLibclaudepending5/6 Param Surface as strategy registry: /param-surface/ index + per-strategy grid pages
106 dAtlasuserpending6/6 indicator_alerts DDL audit: dump schema, commit as ground truth
105 dMultiTraderuserpendingUSER: 3 steps to finish journal persistence
104 dAtlasclaudependingDATA FRESHNESS: audit why newest trades aren't flowing into DB + Param Surface
104 dAtlasclaudependingFix pipeline not running for 2026-04-17 (subset of freshness audit)
104 dAtlasclaudependingFix scorecard manifest missing 4/1-4/13 days (likely same root cause)
101 dMultiTraderuserpendingUSER: run scientific_layer migrations 001–009 via psql tunnel
86 dAcct-v2jimbopendingBUG: Risk Bump Progress bar does not respect the selected date range (always shows 5-day rolling, even when range = full month)
Showing the 12 oldest of 39. Full list in the PM Cockpit.

Freshness alerts every monitor checked, including the monitors themselves

MonitorLast generated (ET)AgeVerdict
/system/freshness.json 2026-05-07 9:38 PM86 d DEAD. Never cron’d — the ticket to schedule update_freshness.py is itself 87 days stale and sits in the blocker table above. Its last snapshot still lists live_classifier and nightly_pm as stale; those two flags are 86 days old and mean nothing today.
/system/app-status.json 2026-07-08 12:01 PM24 d DEAD. The designated replacement for freshness.json, and it stopped too. Its last recorded verdict was overall RED — playbooks 38.7 h (threshold 30 h) and param_surface_db 340.7 h. Those numbers are 24 days old.
atlas-flow-full 2026-07-31 8:30 PM18 h GREEN — 13/13 PASS, and the only monitor still alive. It has not run today (it is a nightly), so 18 h is expected on a Saturday rather than a failure. This is the one to promote to canonical.
multi-tf daemon 2026-07-31 8:00 PM— Running away. Friday’s nightly PM caught cycles #859–#869 rebuilding an ~800 KB page every ~60 s until 8:00 PM ET — four hours past the close — with byte-identical output. It is a no-op rewrite over WebDAV. The run_if_stale.sh guard PVP already uses is the fix.

Today’s standup focus top 3 open P0/P1 by priority × age

1
Atlas 1/6 Data-freshness dashboard: /system/freshness.json + banner on hub
P0 · pending · owner claude · 106 days old · score 212
Every freshness monitor on the property is dead or asleep — this is the ticket that owns the problem. freshness.json has not regenerated in 86 days; the app-status replacement is 24 days cold. The brief cannot tell you what is stale because the thing that measures staleness is itself stale.
2
Atlas 2/6 Pipeline orchestrator: run_nightly.py calls every generator, logs to one place, fails loud
P0 · pending · owner claude · 106 days old · score 212
Today is the failure mode this ticket exists to kill. The HOME box slept, woke, and dumped 14 crons in a 17-minute burst (2:10–2:27 PM ET) — nightly PM, all six MGP editions and this brief all fired at once. One orchestrator with one log and a loud failure ends the silent-miss pattern that has now recurred six times.
3
AMS 4/6 Identity function: resolve_user(account_id) -> (user_id, trader_name)
P0 · pending · owner claude · 106 days old · score 212
Named critical-path dependency for AMS Phase 1. resolve_user(account_id) is what the multi-trader scorecards, the identity port and the DB writes all sit on. It has been open since the Phase 0 freeze.
How these were picked. Score = priority weight (P0 = 2, P1 = 1) × age in days. Six tickets tie at the top on 212 — the whole 1/6–6/6 Phase 0 foundation set, all created 106 days ago and none started. The tie was broken by how much each one unblocks. The three not shown: 3/6 main.daily_summary SQL view and 5/6 Param Surface as strategy registry both wait on the migrations below, and 6/6 indicator_alerts DDL audit is user-owned and already sits at the top of the stale-blocker table.

Critical path · next step first unblockable item · /system/sprint-plan.html

1
MultiTrader James runs the scientific-layer migrations 001–009 via psql tunnel
P0 · pending · owner user · 101 days old · estimated 15 minutes of actual work
Unchanged as step 1 of the critical path since the chain was first drawn. Nothing downstream of it can start. Running it retires 4 user-owned P0s and unblocks p0-daily-summary-view, the rule-based auto-classifier path, and identity_fn DB writes — which is focus item 3 above. This is the single highest-leverage 15 minutes available anywhere on the board, and it has been available for 101 days.
Steps 2–4 behind it: Jim ships the two Acct-v2 date-range bugs (both P0, 80 and 86 d) → Atlas ships the p0-freshness-dashboard MVP → Roberto ships the classifier refresh cron. Step 2 is independent of step 1 and could run in parallel today.