docs: coordination-chat handover for the RAG improvement programme - #2024
Conversation
Adds docs/rag-improvement/COORDINATION.md — the coordinator-side layer over the worker HANDOVER: programme history (PRs #1895/#1899/#1908), the coordinator role definition, the wave dispatch plan and per-packet model/effort/plan-mode settings, the chained worker-prompt handoff pattern, the babysit playbook learned driving #1899 to merge, the standing-approvals map, and a paste-ready coordination-chat bootstrap prompt. Links it from the docs catalogue and HANDOVER, and corrects HANDOVER's now-verifiable status rows (#1908 merged). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UfJeqoq59y5xKmgRGHjqyA
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01UfJeqoq59y5xKmgRGHjqyA
|
Warning Review limit reachedYou’ve reached a temporary PR review limit under our Fair Usage Limits Policy. Next review available in: 3 minutes Limit details: You’ve used all 1 included review currently available under your plan. You completed 94 included PR reviews in the past 7 days; at that activity level, included reviews refill at 1 review per hour. Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (12)
Comment |
|
This pull request has been ignored for the connected project Preview Branches by Supabase. |
…state Resolve the HANDOVER.md status-table conflict against main after PR #2022 (S1, squash 2bd146e) and PR #2023 (#212 T3, squash 440a34f) merged. - HANDOVER §1 snapshot re-dated to 2026-08-17: S1 landed (rung 1, canary pair green), residuals R1/R2/R3 recorded, owner decisions (R1 before S2; governance Option B), #212 sibling stream cross-referenced. - HANDOVER §2 rows: S1 merged; new S1b (R1 routing), S1c (R2+R3), G1 (Option B), #212 T1-T3 / T4 rows; §3 packet bodies and §7 prompts for S1b, S1c, G1. - COORDINATION §3 wave table re-cut (Wave 0: D1 + S4 + S1b + T4; C1 canary done), per-task model/effort table, §7 current state with the canary evidence and the live board link. - clinical-hazard-analysis H5a: partial mitigation noted; document-summary residual and the Option B decision recorded. - Six immutable inbox requests: R1 (P1), R2+R3 (P2), Sentry search error (P2), source_metadata pin (P3), eval-canary post-merge-only note (P3), G1 implementation (P3). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
… ledger (#2045) * docs(issues): reconcile 28 inbox requests into the outstanding-issues ledger Redo of the earlier reconcile: main gained 6 more pending inbox requests (3 cancellations, 1 update, 1 update, 1 done) after this branch's original base, which check:ledger-write-discipline correctly rejected as a partial transaction. Restarted from fresh origin/main and reconciled the complete current pending batch (28 requests) in one transaction, per PRs #2023/#2024 and everything that landed on main since. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2BUayF2Z9DAxvmL5XmLcC * docs(review): record redone ledger-reconcile review Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01D2BUayF2Z9DAxvmL5XmLcC --------- Co-authored-by: Claude <noreply@anthropic.com>
2 participants
Summary
docs/rag-improvement/COORDINATION.md: the coordinator-side layer completing the programme's three-file structure (README = design authority, HANDOVER = worker packets, COORDINATION = how the owner's coordination chat runs the programme). It records the programme history to date, defines the coordinator role versus worker sessions, the wave dispatch plan, per-task model / reasoning-effort / plan-mode settings, the chained worker-prompt handoff pattern, the babysit playbook, the standing-approvals map, and a paste-ready coordination-chat bootstrap prompt.HANDOVER.mdstatus-table conflict againstmainafter PR fix(rag): stop verification rejecting verbatim-faithful answers (#231 S1) #2022 (S1, squash2bd146eed) and PR harden(api): validate inbound DB and RPC payloads against Zod row contracts #2023 (fix(supabase): make GUC-set migration hosted-safe; neutralize obsolete N+1 migration #212 T3, squash440a34f71) merged, and brings the programme docs to truth:HANDOVER.md§1 snapshot re-dated to 2026-08-17 — S1 landed at ladder rung 1 with its canary pair green (baseline run 31964560921 on8f8d111ab→ post run 32025082010 on2bd146eed: document/content recall 1.0/1.0, zero per-case rr regressions, answer gate 45/45); residuals R1/R2/R3 recorded; owner decisions (R1 before S2; governance Option B); the#212sibling stream cross-referenced.HANDOVER.md§2 rows: S1 merged; new rows S1b (R1 pre-deadline strong routing), S1c (R2+R3 claim-support strictness), G1 (Option B provenance tag),#212T1–T3 and T4; §3 packet bodies and §7 paste-ready prompts for S1b, S1c and G1.COORDINATION.md§3 wave table re-cut (Wave 0 = D1 + S4 + S1b + T4, C1 canary done; Wave 1 = R0 reconcile + G1 + S1c; Wave 2 = S2 + S5 + S6; Wave 3 = S3), one-chat-per-task model/effort table, §7 current state with the canary evidence and the live board link.docs/clinical-hazard-analysis.mdH5a: notes the partial mitigation now inderiveConfidence(strongestNonSynthetic), names the untagged document-summary path as the live residual, and records the owner's Option B decision with Option A rejected.docs/outstanding-issues-inbox/(no canonical ledger edit): R1 unbudgeted strong escalation (P1), R2+R3 claim-support strictness (P2), recurring SentryUnhandled server request erroron the search routes (P2),source_metadatadata-backed pin (P3), eval-canary post-merge-only protocol note (P3), G1 implementation task (P3). Apply withnpm run issues:reconcileafter this lands.docs/README.mdandHANDOVER.md.RAG impact: no retrieval behaviour change — documentation and ledger inbox requests only; no protected RAG surface is modified.
Verification
npm run verify:pr-local -- --files <all changed paths>on head8e7885913— docs-focused scope:completed: check:runtime, check:installed-lock-parity, format:changed, sitemap:check, docs:check-index, docs:check-inventory, docs:check-scripts, docs:check-links, check:branch-review-ledger, check:outstanding-issues, check:ledger-write-discipline — failed: (none). Decisive lines:Ledger inbox check passed: 22 pending request(s), 189 applied.·Ledger write discipline passed for bb3d4e7ab0dc..HEAD.·docs link check passed: 1794 repo path references resolve.git diff --stat origin/main...HEAD: 11 files, all underdocs/— nosrc/,tests/,scripts/change.main, not this branch.)Risk and rollout
Notes
docs/branch-review-records/446cd9ee…record.md(supersedes the earlier record for this branch).Generated by Claude Code; synced 2026-08-17 by the RAG programme coordinator chat.