Daily Progress — 2026-05-01

Friday · Generated by nightly-project-manager · ← All Todos · ← Apr 30
4
Shipped
4
Moved Forward
8
Blocked

What Shipped

What Moved Forward

What's Blocked

TaskOwnerBlocker
run-scientific-layer-migrations 001–009 (P0)user10 days running. Auto-classifier rule-based path now genuinely needs migration 001 columns to upgrade from heuristic to true rule-based. ~15 min via set_db.sh start tunnel.
journal-user-steps (P0)userRun add_scorecard_columns.sql, confirm user_id=2, click Save
p0-indicator-alerts-ddl (P0)userpsql -c '\d alerts_tradingview.indicator_alerts' dump needed
p0-daily-summary-view (3/6)userMigration 004 staged but unrun
p0-freshness-dashboard (1/6)claudeToday's missing recap MGP is the third silent-regression instance in 5 days. The case for shipping this just got stronger again.
data-freshness-audit (P0)claudeNeeds DB access / SSH tunnel
pm-activity-scanner-deploy (P1)userImport task_scheduler_import.xml on DESKTOP-AKU8I62
MGP recap edition for 5/1 missingclaudeNo 2026-05-01-recap.html on disk; recap notify did not fire. 3rd recap failure in 5 days (4/28, 4/29 missed too).

Tomorrow's Priority

  1. Catalog refactor to one shared JSON — carried over from yesterday. _scientific_overlay.js, /strategies/, /sizing/, playbooks.json all inline the same 16-strategy catalog. Now that the auto-classifier is end-to-end correct, drift between surfaces is the next failure mode. Extract to /strategies/catalog.json and have all 4 surfaces fetch from one source.
  2. Investigate the recurring recap-MGP failure — three out of the last five trading days (4/28, 4/29, 5/1) had no recap edition. Check gen_mgp_intraday.py recap branch + scheduled-task firing time + last successful recap log. Either crash-without-alert or scheduled-task off.
  3. p0-freshness-dashboard (1/6) — fully Claude-owned. Each silent recap miss is one more datapoint. Build /system/freshness.json + stale-banner; this would have caught all three misses within an hour.

Files Touched

Generated by nightly-project-manager · 4 shipped, 4 moved forward, 8 blocked · Top priority tomorrow: catalog refactor to one shared JSON + investigate recurring recap-MGP failure