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>
70 lines
3.4 KiB
JSON
70 lines
3.4 KiB
JSON
[
|
||
{
|
||
"date": "2026-05-07",
|
||
"client": "Riverfront Steel",
|
||
"cities": "Columbus",
|
||
"states": "OH",
|
||
"events_total": 4,
|
||
"events_ok": 3,
|
||
"checkpoint_count": 2,
|
||
"model": "gpt-oss:20b",
|
||
"cloud": false,
|
||
"lesson": "** \nWhen calling `hybrid_search`, first validate the executor’s JSON output against the expected schema and wrap the call in error‑handling logic to catch parse failures. If a parse error occurs, retry or fall back to a safe default before proceeding. Before assigning a new shift to any candidate, query their current schedule to confirm no overlap—especially for high‑demand staff such as Jack Kim. This double‑check prevents double‑booking and keeps pool counts accurate.",
|
||
"checkpoints": [
|
||
{
|
||
"after": "17:30",
|
||
"risk": "JSON parse failure in tool call",
|
||
"hint": "Validate executor output format; add error handling around hybrid_search; confirm JSON schema before using results."
|
||
},
|
||
{
|
||
"after": "15:30",
|
||
"risk": "Jack Kim double-book risk",
|
||
"hint": "Verify Jack Kim's 12:00 slot before assigning new shift to avoid overlap."
|
||
}
|
||
],
|
||
"created_at": "2026-04-21T02:09:33.271Z",
|
||
"file": "2026-05-07_Riverfront_Steel_1776737373271.json"
|
||
},
|
||
{
|
||
"date": "2026-04-28",
|
||
"client": "Riverfront Steel",
|
||
"cities": "Toledo",
|
||
"states": "OH",
|
||
"events_total": 3,
|
||
"events_ok": 2,
|
||
"checkpoint_count": 1,
|
||
"model": "gpt-oss:20b",
|
||
"cloud": false,
|
||
"lesson": "** \nBefore scheduling a Packer in Toledo, confirm the tool calibration and verify the location accuracy to mitigate drift risk. Update the pool status immediately after baseline and recurring fills so the Packer has accurate data. Schedule the Packer only after the pool is refreshed to avoid misplacement failures. Double‑check the Packer’s tool settings in the system before the shift starts.",
|
||
"checkpoints": [
|
||
{
|
||
"after": "08:30",
|
||
"risk": "Packer Toledo drift risk",
|
||
"hint": "Check Packer's tool calibration and location accuracy in Toledo before scheduling to avoid drift errors."
|
||
}
|
||
],
|
||
"created_at": "2026-04-21T01:52:09.394Z",
|
||
"file": "2026-04-28_Riverfront_Steel_1776736329394.json"
|
||
},
|
||
{
|
||
"date": "2026-04-26",
|
||
"client": "Apex Warehouse",
|
||
"cities": "Cleveland",
|
||
"states": "OH",
|
||
"events_total": 3,
|
||
"events_ok": 2,
|
||
"checkpoint_count": 1,
|
||
"model": "gpt-oss:20b",
|
||
"cloud": false,
|
||
"lesson": "** \nWhen an expansion event is scheduled before a recurring event, always pre‑fetch the availability of every candidate slated for the recurring slot (e.g., Anna Rivera, Jack K. Ramirez, Thomas J. Lewis, Brenda A. Cox, Lisa O. Diaz) at the overlapping time. Verify that none of these candidates are already assigned to the expansion or any other role before committing them to the recurring slot. Adjust the expansion or recurring schedule if any overlap is detected, and keep the pool size and turns in mind to avoid overloading the same candidates. This prevents double‑booking and ensures a smooth fill.",
|
||
"checkpoints": [
|
||
{
|
||
"after": "14:30",
|
||
"risk": "Double-booking at 15:30",
|
||
"hint": "Verify 15:30 slot for Anna Rivera, Jack K. Ramirez, Thomas J. Lewis, Brenda A. Cox, Lisa O. Diaz before assigning new roles."
|
||
}
|
||
],
|
||
"created_at": "2026-04-21T01:48:13.293Z",
|
||
"file": "2026-04-26_Apex_Warehouse_1776736093293.json"
|
||
}
|
||
] |