Repository navigation
test: key the golden extents and pipeline exports per entry, not by a shared total - #37679
Conversation
UX Review (Opus 5.5) — ⏭️ skippedRevision |
First Principles Review (Opus 5.5) — ⏭️ skippedRevision |
Design Review (Opus 5.5) — ✅ PASSDesign-level review of Design-Verdict: PASS No design-level concerns. Each per-entry record is at least as strict as the shared total it replaces. [DESIGN-REVIEWED] 02d6cf9 Description read[DESCRIPTION-READ] 3a10cc3fe83532ff6297096f50f69ed735bf7e48d8764fb293fdcccb415fe6b8 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 |
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: |
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: |
|
Rebased: main moved in 4 commits (d470c46 -> 57eba16), one of which re-pinned test/test_memory_v1_golden.py's EXPECTED_TOTAL to 8776 and two block extents ([CRITICAL RULES 5057, [Learned corrections 374) while this PR removes the total. Resolved by keeping main's extents and the removal; the per-block extents and the sum == payload length check still decide. Re-ran test_memory_v1_golden.py and test_channel_turn_structure.py on the new base: 43 passed. New head: 02d6cf9. |
80334b7 to
02d6cf9
Compare
Code change approver recommendation: ApproveWhy: the decider estimates a 5% chance that this change introduces a defect, within the 12% limit for approving on its own, and every check passes. Test-only change. In the memory golden test, the removed EXPECTED_TOTAL (8776) equals the sum of the exact per-block EXPECTED_EXTENTS, and the payload coverage check stays, so nothing is loosened. In the export test, the shared count is replaced by an exact set of 31 names, one per line, checked in both directions, which is stricter than the count. CI is green on this head, so the set matches the current imports. No other code reads the removed constants. An automated recommendation, not a review: a maintainer still decides. It approves on its own only when the estimated chance of a defect is at most 12% and the change has at most 400 changed lines. Decider v3c (217fe607) · commit |
KiroCrewCaptain
left a comment
There was a problem hiding this comment.
Tier 1 auto-approve: test (2 files). Criteria: no conflict, no requested changes, security path denylist clean, design-doc gate clean, security checklist all-NO, no reviewer-lane findings, readiness signal fresh for this head SHA, declared category matches the diff. Category: test-only change that replaces a shared export count and a pinned golden total with per-entry sets, touching only test/ files.
Problem / Motivation
Goal: The memory golden and the channel-turn export tests stay green on
mainwhen two pull requests that were each green on the same base both merge.Each stored one shared total beside or instead of a per-entry record. Two PRs that each change a different entry rewrite the total to the same value, git merges it cleanly, and
mainholds a different total than the number says.EXPECTED_TOTAL = 8003intest/test_memory_v1_golden.pywas re-pinned 4 times in 30 days._EXPORT_BUDGET = 31intest/test_channel_turn_structure.pysat exactly at its measured value, so any two PRs that each add an import collide.Why it matters
A red
mainfails every open PR's CI until someone re-pins the number.Not a goal
test/test_review_repair_routing_skill.pyandPROMPT_BYTE_CEILINGSare real ceilings, where a red after a merge means the file really grew past its budget.What changed (motivation → approach → change)
test/test_memory_v1_golden.pyEXPECTED_TOTALduplicates the sum of the exact per-blockEXPECTED_EXTENTStest/test_channel_turn_structure.py_EXPORT_BUDGET = 31, one shared count of imported pipeline names_EXPORTED_NAMES, onemodule.nameper line. A new import fails naming it; a retired one fails asking to delete its lineThe rule is the one in
docs/system-specs/common/testing-conventions.md§ Merge-safe ratchets and counts (PR #37660).Backwards compatibility
Compatible: test files only.
Removes nothing: the golden total is implied by the exact block lengths and the coverage check that stay; the export count is replaced by a stricter exact set.
Tests
Two-branch merge simulation (scratch clone; A and B cut from one base, each makes one change of the same kind and updates the record exactly as the test asks, then
git merge):origin/mainassert 8017 == 8010242vs235)33 names are imported ... over the budget of 32new names imported from the pipeline's modules: [...]On this branch rebased onto current
main: both changed files plustest/test_flake_pattern_ratchet.py, 111 passed; black, isort and flake8 clean.Proof: not provable -- test-only change
Manual verification
N/A — tests only; the simulation above is the behavioral check.
Related Issues
Part of #37656
Checklist