cat-mode: lead the multi-emitter backfill done-gate with the general lesson.

[EdbertChan]

Depends-On: #1311

Summary

This updates a working-style lesson for agents who are asked to audit a broad backfill. It tells them to prove every writer and source, not just quote the request.

The problem was that the rule led with example phrases. That made the lesson sound tied to those phrases instead of the wider failure.

The fix leads with the general backfill requirement, then keeps the old phrases as examples. The nearby checks now lock that wording.

Review Claim

Review that the backfill lesson now leads with the general requirement while preserving the examples that trigger it.

Review Lane

behavior

Review Unit

corpus-lesson

Safety Invariant

This slice changes lesson wording and regression locks only. It does not alter runtime code.

Assumption: this invariant is unconfirmed because this is a headless CI repair.

Slice Rationale

This stays in the lesson slice so the reviewer can approve the wording rule and its checks together.

Non-goals

• No runtime behavior change.
• No new lesson trigger.
• No publishing flow change.

Measured

none: Wording-only lesson slice; there is no runtime metric to compare.

Test Plan

Test Plan

• node engine/skills/draft-pr/scripts/validate-pr-body.mjs --body "$BODY" --changed-files-file <(git diff --name-only origin/stack/EdbertChan/automate/dashboard-health-receipt/cat-mode-align-publishing-analytics-fixture--debc0f0b...origin/stack/EdbertChan/automate/dashboard-health-receipt/cat-mode-lead-multi-emitter-backfill-done-gate--aa84683f)
• python3 engine/skills/make-pr/scripts/preflight.py --base origin/stack/EdbertChan/automate/dashboard-health-receipt/cat-mode-align-publishing-analytics-fixture--debc0f0b --commit-only --dry-run

Revert Plan

Revert Plan

• Safe to revert? Yes
• Revert command: git revert 07f048683f699506e7f071d9528e23a849c6225c
• Post-revert steps: None
• Data migration? No

---

Note

Low Risk
Wording-only changes to cat-mode skill docs and unittest fixtures; no runtime or product behavior changes.

Overview
cat-mode publishing-analytics rules are tightened so agents fix data at emit/resolve/backfill instead of patching charts, and prove coverage before claiming “done.”

Emit/resolve: Blank or synthetic model is an emit or resolve defect (not emit-only). Chart filters that exclude or include blanks are not fixes; prove the live resolve path first.

Backfill done-gate: “Full backfill” is satisfied only with a harness×source matrix plus fill table, with example phrases (full backfill, all harnesses) secondary to that general requirement.

New gates: Require a health receipt (harness×action, filled_model_rate, volume/fill/resolve/filter ownership) before narrating by-model health; split those four before L3 or insight/notebook PATCH when a harness is “missing” or real-model is zero; when /automate-me (or similar) lands mid thrash, lock the class catch before more analytics PRs.

references/verify.md mirrors the same text; new fires_/stays_silent_ fixtures and expanded TestPublishingAnalyticsDefaults / TestAnalyticsThrashCheckpoints lock the wording.

^{Reviewed by Cursor Bugbot for commit 30840ff. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: 5476db25-4bba-4cb4-9d61-b54e687e53f6)

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: d5c1e6d3-b0ee-4268-8912-c65d22301187)

[mergify]

Queued — the merge queue status continues in this comment ↓.

[EdbertChan]

@Mergifyio queue

[mergify]

Merge Queue Status

• ✅ Entered queue — 2026-10-13 08:22 UTC · Rule: admin-bypass · triggered by @EdbertChan with the @mergifyio queue command
• ✅ Checks passed · on draft merge queue: checking #1312 on main (4c092f7) #1394
• ✅ Merged — 2026-10-13 09:09 UTC · at ff82cf89b9897a903904fa303403ba1e0f4e1ad6 · squash

This pull request spent 46 minutes 45 seconds in the queue, including 46 minutes 22 seconds running CI.

Required conditions to merge

• check-success = lint
• check-success = test
• check-success = validate
• depends-on = EdbertChan/catstack#1311 [⛓️ cat-mode: align publishing-analytics fixture asserts with revised example phrasing. #1311]
  • cat-mode: lead the multi-emitter backfill done-gate with the general lesson. #1312