[auto] cat-mode: fold direct answers and cross-session rules, drop dated provenance

[EdbertChan]

Revise three contradicted lines in place (own-repo destructive action
executes after a verified list; hedges auto-run prove-it; waits are
sleep/wake with a clock-time ETA, never a poll loop) and fold nine
high-confidence mined rules into the sections that already fit them.
Strip every "Found via"/dated provenance clause per automate-me's
guardrail so the file stays under the 300-line lock (295 -> 300).

Tests: tests/test_cat_mode.py locks each revised or new line and asserts
no dated provenance remains.

Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com
Claude-Session: https://claude.ai/code/session_019CoNSA1EKhBHfVUX7feoKR

---

Note

Low Risk
Changes are skill prose and unittest guards only; no product runtime or auth/data paths are modified.

Overview
Updates cat-mode so three previously contradictory behaviors are stated plainly: durable waits use sleep/wake with a clock-time ETA (explicitly not poll/resume), hedges auto-run prove-it in the same turn, and in the user's own repo "I am in control, just do it" means show a verified impact list then execute without a consent loop.

The same edit folds mined cross-session rules into existing sections—"land" through CI/conflicts/deploy, resume-in-place after accidental interrupts, design proposals after history scan, worker-owned work queued not hand-fixed, new rules naming prior art, recurring interventions routed to hooks, one-off scripts folded into a single entry point, real-surface proof and [[principle-explicit-errors]]—and removes all "Found via …" / dated provenance while staying at the 300-line cap.

tests/test_cat_mode.py updates the wait/wake assertion, adds TestCatModeDirectAnswers to lock each new or revised phrase, and asserts no dated provenance remains in the skill file.

^{Reviewed by Cursor Bugbot for commit e364344. 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: serverGenReqId_c75c4050-ba52-48d3-ad19-fd5ec24acd8f)

[EdbertChan]

This pull request is part of a Mergify stack:

#	Pull Request	Link
1	[auto] cat-mode: fold direct answers and cross-session rules, drop dated provenance	#244	👈
2	test: drop two comment lines (no-comments gate)	#245

[EdbertChan]

Revision history

#	Type	Changes	Reason	Date
1	initial	024f875		2026-09-07 08:06 UTC
2	content	024f875 → e364344 (raw)		2026-09-07 08:06 UTC

[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: serverGenReqId_6e18ad6d-bbf4-4434-880d-dd6b66cd69b3)

[EdbertChan]

@Mergifyio queue

[mergify]

Merge Queue Status

• ✅ Entered queue — 2026-09-07 16:52 UTC · Rule: admin-bypass · triggered by @EdbertChan with the @mergifyio queue command
• ✅ Checks passed · on draft merge queue: checking #244 on main (7c75dac) #258
• ✅ Merged — 2026-09-07 16:56 UTC · at d2754577e21cc687d210cbf33532ac15b0aa26e8 · squash

This pull request spent 3 minutes 52 seconds in the queue, including 3 minutes 10 seconds running CI.

Required conditions to merge

• check-success = lint
• check-success = test