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>
5.4 KiB
5.4 KiB
Scenario retrospective — Riverfront Steel, 2026-04-21
Executor: qwen3.5:latest Reviewer: qwen3:latest Draft: qwen2.5:latest Overview(T3): gpt-oss:20b
Prior lessons loaded into executor context: 1 (from 2026-04-21)
Events
| At | Kind | Role / Count | Pool | Fills | Turns | Dur(s) | Cites | Gaps |
|---|---|---|---|---|---|---|---|---|
| 08:00 | baseline_fill | Warehouse Associate × 3 | 770 | ✓ 3 | 5 | 37.1 | 0 | 2 |
| 10:30 | recurring | Machine Operator × 2 | 997 | ✓ 2 | 2 | 14.6 | 0 | 2 |
| 12:15 | expansion | Forklift Operator × 5 | 1184 | ✓ 5 | 6 | 50.0 | 0 | 5 |
| 14:00 | emergency | Loader × 4 | - | ✗ 0 | 0 | 47.1 | 0 | 1 |
| 15:45 | misplacement | Warehouse Associate × 1 | 770 | ✓ 1 | 3 | 18.5 | 0 | 1 |
Final roster
| Worker | Booked | Role | City, ST | Status |
|---|---|---|---|---|
| undefined Patrick Ross | 08:00 | Warehouse Associate | Toledo, OH | no_show |
| undefined Olivia Y. Howard | 08:00 | Warehouse Associate | Toledo, OH | confirmed |
| undefined Deborah X. Sanchez | 08:00 | Warehouse Associate | Toledo, OH | confirmed |
| undefined Kimberly G. Thomas | 10:30 | Machine Operator | Toledo, OH | confirmed |
| undefined Kathleen O. Ortiz | 10:30 | Machine Operator | Toledo, OH | confirmed |
| undefined Matthew P. Garcia | 12:15 | Forklift Operator | Toledo, OH | confirmed |
| undefined Maria K. Cruz | 12:15 | Forklift Operator | Toledo, OH | confirmed |
| undefined Nancy W. Ward | 12:15 | Forklift Operator | Toledo, OH | confirmed |
| undefined Charles T. Walker | 12:15 | Forklift Operator | Toledo, OH | confirmed |
| undefined Rachel Turner | 12:15 | Forklift Operator | Toledo, OH | confirmed |
| undefined Ryan Hughes | 15:45 | Warehouse Associate | Toledo, OH | confirmed |
Gap signals
double_book
- 08:00 — undefined Olivia Y. Howard already booked for 08:00
- 08:00 — undefined Deborah X. Sanchez already booked for 08:00
- 10:30 — undefined Kimberly G. Thomas already booked for 08:00
- 10:30 — undefined Kathleen O. Ortiz already booked for 08:00
- 12:15 — undefined Matthew P. Garcia already booked for 08:00
- 12:15 — undefined Maria K. Cruz already booked for 08:00
- 12:15 — undefined Nancy W. Ward already booked for 08:00
- 12:15 — undefined Charles T. Walker already booked for 08:00
- 12:15 — undefined Rachel Turner already booked for 08:00
- 15:45 — undefined Ryan Hughes already booked for 08:00
drift_or_tool
- 14:00 — invalid JSON from executor: JSON Parse error: Expected '}' | raw: {"kind":"propose_done","fills":[{"candidate_id":"W500K-12325","name":"Raj Torres","reason":"Top-ranked Loader in Toledo, OH with high availability (score 0.72) and relevant skills (SAP, hazmat)."},{"candidate_id":"W500K-16975","name":"Brian X. Price","reason":"Second-ranked Loader in Toledo, OH with
fairness
- cross-event — Patrick Ross (undefined) booked 10 times today
write_through_audit
- post-run — playbook_memory has 1477 entries (ran 5 events, expected ≥ 4 new entries from this run)
Workers touched across the week
12 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 |
|---|---|---|---|
| 7079 | Patrick Ross | 08:00 baseline_fill | booked |
| 48488 | Olivia Y. Howard | 08:00 baseline_fill | booked |
| 39023 | Deborah X. Sanchez | 08:00 baseline_fill | booked |
| W500K-48548 | Kimberly G. Thomas | 10:30 recurring | booked |
| W500K-25702 | Kathleen O. Ortiz | 10:30 recurring | booked |
| 22375 | Matthew P. Garcia | 12:15 expansion | booked |
| 19588 | Maria K. Cruz | 12:15 expansion | booked |
| 28024 | Nancy W. Ward | 12:15 expansion | booked |
| 17543 | Charles T. Walker | 12:15 expansion | booked |
| 9076 | Rachel Turner | 12:15 expansion | booked |
| 11915 | Ryan Hughes | 15:45 misplacement | booked |
| undefined | Patrick Ross | 08:00 | no_show |
Discovered patterns (meta-index)
What the system identified across semantically-similar past fills as each event ran:
- 08:00 baseline_fill (Warehouse Associate): Across 25 similar past playbooks (11 workers examined) · recurring certifications: OSHA-10 (64%), Forklift (45%) · recurring skills: overhead crane (45%) · archetype mostly: communicator · reliability median 0.96 (range 0.75–1.00)
- 10:30 recurring (Machine Operator): Across 25 similar past playbooks (11 workers examined) · recurring certifications: OSHA-10 (64%), Forklift (45%) · recurring skills: overhead crane (45%) · archetype mostly: communicator · reliability median 0.96 (range 0.75–1.00)
- 12:15 expansion (Forklift Operator): Across 25 similar past playbooks (12 workers examined) · recurring certifications: OSHA-10 (67%) · recurring skills: mill (42%), overhead crane (42%) · archetype mostly: communicator · reliability median 0.96 (range 0.59–1.00)
- 14:00 emergency (Loader): —
- 15:45 misplacement (Warehouse Associate): Across 25 similar past playbooks (12 workers examined) · recurring certifications: OSHA-10 (67%) · recurring skills: overhead crane (42%), mill (42%) · archetype mostly: communicator · reliability median 0.96 (range 0.59–1.00)
Narrative
- 4/5 events reached consensus.
- Final roster: 11 bookings across 1 distinct workers.
- Workers touched (booked, failed, or otherwise decided): 12.
- Playbook citations across the day: 0 (proof the feedback loop fired across events).
- Dropped events: 14:00 emergency.