41b0a99ed2
Surfaced by today's untracked-files audit. None of these are accidents —
multiple are referenced by name in CLAUDE.md and memory files but were
never added.
Categories:
- docs/PHASE_AUDIT_GUIDE.md (106 LOC) — Claude Code phase audit guidance
- ops/systemd/lakehouse-langfuse-bridge.service — Langfuse bridge unit
- package.json — top-level npm manifest
- scripts/e2e_pipeline_check.sh + production_smoke.sh — real test scripts
- reports/kimi/audit-last-week*.md — the "Two reports live" CLAUDE.md cites
- tests/multi-agent/scenarios/ — 44 staffing scenarios (cutover decision A)
- tests/multi-agent/playbooks/ — 102 playbook records
- tests/battery/, tests/agent_test/PRD.md, tests/real-world/* — real tests
- sidecar/sidecar/{lab_ui,pipeline_lab}.py — 888 LOC dev-only UIs that
remain in service post-sidecar-drop (commit ba928b1 explicitly kept them)
Sensitivity check: scenarios use synthetic company names ("Heritage Foods",
"Cornerstone Fabrication"); audit reports describe code findings only;
no PII or secrets surfaced.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
4.4 KiB
4.4 KiB
Scenario retrospective — Midway Distribution, 2026-05-05
Executor: qwen3.5:latest Reviewer: qwen3:latest Draft: qwen2.5:latest Overview(T3): gpt-oss:20b
Prior lessons loaded into executor context: 2 (from 2026-04-29, 2026-04-21)
Events
| At | Kind | Role / Count | Pool | Fills | Turns | Dur(s) | Cites | Gaps |
|---|---|---|---|---|---|---|---|---|
| 16:00 | baseline_fill | Assembler × 2 | 838 | ✓ 2 | 3 | 26.0 | 0 | 2 |
| 17:30 | recurring | Loader × 2 | 620 | ✓ 2 | 2 | 11.8 | 0 | 3 |
| 10:30 | expansion | Packer × 5 | - | ✗ 0 | 0 | 37.9 | 0 | 1 |
| 12:00 | emergency | Loader × 4 | 431 | ✓ 4 | 4 | 35.0 | 0 | 5 |
Final roster
| Worker | Booked | Role | City, ST | Status |
|---|---|---|---|---|
| undefined Zachary Taylor | 16:00 | Assembler | Indianapolis, IN | confirmed |
| undefined Adam D. Watson | 16:00 | Assembler | Indianapolis, IN | confirmed |
| undefined Sharon B. Torres | 17:30 | Loader | Indianapolis, IN | confirmed |
| undefined Kimberly Murphy | 17:30 | Loader | Indianapolis, IN | confirmed |
| undefined Sandra Gray | 12:00 | Loader | Indianapolis, IN | confirmed |
| undefined Mark Lopez | 12:00 | Loader | Indianapolis, IN | confirmed |
| undefined Helen Z. Cooper | 12:00 | Loader | Indianapolis, IN | confirmed |
| undefined Ryan Z. Torres | 12:00 | Loader | Indianapolis, IN | confirmed |
Gap signals
double_book
- 16:00 — undefined Adam D. Watson already booked for 16:00
- 17:30 — undefined Sharon B. Torres already booked for 16:00
- 17:30 — undefined Kimberly Murphy already booked for 16:00
- 12:00 — undefined Sandra Gray already booked for 16:00
- 12:00 — undefined Mark Lopez already booked for 16:00
- 12:00 — undefined Helen Z. Cooper already booked for 16:00
- 12:00 — undefined Ryan Z. Torres already booked for 16:00
artifact
- 16:00 — undefined is not an object (evaluating 'f.reason.slice')
- 17:30 — undefined is not an object (evaluating 'f.reason.slice')
- 12:00 — undefined is not an object (evaluating 'f.reason.slice')
drift_or_tool
- 10:30 — invalid JSON from executor: JSON Parse error: Unable to parse JSON string | raw: {"kind":"plan","steps":["Analyze prior search results: The hybrid_search tool returned 0 SQL matches for 'Indianapolis, IN' despite the filter, indicating the vector reranker is returning irrelevant candidates (Memphis, MO, etc.) because no local Packer data exists in the index.","Action: Execute a
fairness
- cross-event — Zachary Taylor (undefined) booked 8 times today
write_through_audit
- post-run — playbook_memory has 1547 entries (ran 4 events, expected ≥ 3 new entries from this run)
Workers touched across the week
8 distinct workers made it through to a decision. Every one is accounted for below — no-shows flagged, rebookings noted, everyone visible.
| Worker ID | Name | Events | Outcome |
|---|---|---|---|
| 12207 | Zachary Taylor | 16:00 baseline_fill | booked |
| 49897 | Adam D. Watson | 16:00 baseline_fill | booked |
| W500K-25481 | Sharon B. Torres | 17:30 recurring | booked |
| W500K-10744 | Kimberly Murphy | 17:30 recurring | booked |
| W500K-11321 | Sandra Gray | 12:00 emergency | booked |
| W500K-7233 | Mark Lopez | 12:00 emergency | booked |
| W500K-23191 | Helen Z. Cooper | 12:00 emergency | booked |
| W500K-26233 | Ryan Z. Torres | 12:00 emergency | booked |
Discovered patterns (meta-index)
What the system identified across semantically-similar past fills as each event ran:
- 16:00 baseline_fill (Assembler): Across 25 similar past playbooks (25 workers examined) · recurring certifications: OSHA-10 (48%) · archetype mostly: communicator · reliability median 0.80 (range 0.59–1.00)
- 17:30 recurring (Loader): Across 25 similar past playbooks (25 workers examined) · recurring certifications: OSHA-10 (48%) · archetype mostly: communicator · reliability median 0.80 (range 0.59–1.00)
- 10:30 expansion (Packer): —
- 12:00 emergency (Loader): Across 25 similar past playbooks (27 workers examined) · recurring certifications: OSHA-10 (52%) · recurring skills: overhead crane (41%) · archetype mostly: communicator · reliability median 0.80 (range 0.19–1.00)
Narrative
- 3/4 events reached consensus.
- Final roster: 8 bookings across 1 distinct workers.
- Workers touched (booked, failed, or otherwise decided): 8.
- Playbook citations across the day: 0 (proof the feedback loop fired across events).
- Dropped events: 10:30 expansion.