Repository navigation
docs(prepare-pr): bring SKILL.md back under its byte ceiling - #16719
bolichen97 wants to merge 1 commit into
Conversation
Design Review (Opus 5.5) — ✅ PASSDesign-level review of Design-Verdict: PASS Nothing to check. [DESIGN-REVIEWED] 00ab9b5 Description read[DESCRIPTION-READ] 48498c81a68413faf32281ee7b378fd2be0df67585fcc75c0ef2ac766987dc73 That sha256 names everything this verdict read: the PR description, plus the 1 evidence file(s) the workflow collected from it and handed this review. It covers that evidence because the media strip replaces every attachment URL with the same placeholder, so an edit swapping one attachment for another would leave the description's captured text identical while what this review judged changed. Recompute it with |
First Principles Review (Opus 5.5) — ✅ PASSPremise-level review of First-Principles-Verdict: PASS The trimmed intro drops "not on every load". Check that the loop's lazy-load guidance for What this change shipsInventory (3 items) — 3 justifiedIntent: get
[FIRST-PRINCIPLES-REVIEWED] 00ab9b5 Description read[DESCRIPTION-READ] 2d59841e749a6a836e4150e5c2c0e8a3081e211fe5c0185d0cd9a414dd0f32fe That sha256 names the PR description this verdict read. Recompute it with |
Opus 5.5 Review — ✅ no blocking findingsReviewed Verdict parsed from the review's SHA-scoped output markers for commit False positive or not applicable? A repository writer can comment: |
|
Closing: #16717 merged and already brings SKILL.md back under its ceiling (main is now 67995 bytes), so this is no longer needed. |
|
Superseded by #16717, which merged the same byte-ceiling fix. |
GPT 6.1 Review — ✅ no blocking findingsGPT 6.1 completed its review of This comment is updated in place on each push. Review detailsNo findings. False positive or not applicable? A repository writer can comment: |
Problem / Motivation
Goal:
prepare-pr/SKILL.mdis back under its 68000-byte ceiling onmain.test_slim_skills_do_not_regrow_duplicate_guidance_or_wire_model_ids[prepare-pr-68000]fails onmain(68077 bytes, 77 over) since #16196. Every open PR inherits the red through its merge ref, in Backend Tests (3.12) shard 3 and Backend Tests (Windows) shard 3.Why it matters
A red test on
mainreds every PR, so no author can tell their own failures from this one.Not a goal
What changed (motivation → approach → change)
77 bytes over → shorten prose that restates itself, leave every pinned phrase alone →
src/kiro_crew/builtin_skills/kirocrew-dev/prepare-pr/SKILL.mdgoes from 68077 to 67984 bytes (−93): the "This file carries only…" paragraph, theaccepted-and-deferredrow (same facts, fewer words), and one clause in thesingle_commitnote. One trim I tried (runtime/provider-reported) is pinned bytest_model_selection_is_discovered_and_service_evidence_is_not_invented, so it stays.Backwards compatibility
No behaviour change; documentation only.
Reader: test/test_review_repair_routing_skill.py:test_slim_skills_do_not_regrow_duplicate_guidance_or_wire_model_ids -- entry: release -- reads the file size; now 67984 ≤ 68000
Reader: test/test_prepare_pr_status.py -- entry: release -- unaffected, passes
Tests
pytest test/test_review_repair_routing_skill.py test/test_prepare_pr_status.py: 229 passed. Onmainthe first file fails withassert 68077 <= 68000.Manual verification
Byte count of the file before and after (68077 → 67984), no CRLF introduced.
Related Issues
Closes #16718
Checklist
feat|fix|docs|style|refactor|perf|test|chore|ci|build|revert: ...)Contribution License Agreement