Repository navigation
Conversation
The results pull request no longer edits ci.yml. A replay-variants job lists the certified variants of data/uk/replay_variants.json in index order (.github/scripts/uk_replay_engine_matrix.py, the standard library under the runner's python3 -I, from a sparse checkout of that script and the index), and replay-engine's matrix is that list. The lister refuses the whole matrix on a record it cannot use. The builder drops add_matrix_variant, its only yaml import and ci.yml from EDITED; apply needs no package in either job; scope refuses any path under .github/. The App token drops permission-workflows, the build job's CONTROL drops pyyaml, and propose explains a refused rebuild over a main that changed a workflow. Finding 4/6 of both reviews of #193. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The refused-rebuild hint now also names a workflow change that main made and then reverted: it checks every commit from the old tip, not just the tips. Variant ids are capped at 120 characters in the lister and the builder alike, so two ids joined by "__" (a release diff's directory) fit one 255-byte path component and the matrix output stays far under GitHub's 1 MB per job. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Default path-limited history prunes a merge to a parent its paths match, so a workflow change and its revert made on a side branch and merged after main moved on were invisible to the hint. --full-history shows them; the test's stand-in for GitHub reads full history too. Answers the review of 34fba89. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The first automated results PR (#195, 6.7.5) failed 30 tests in this file that named 6.5.0 as the newest certified variant or the last committed one: the end-to-end harness certified a synthetic 6.6.0, now refused as older than 6.7.5; the NEWEST/COMMITTED round trip stripped 6.5.0's lines; the paragraph-placement test expected 6.5.0's closing sentence last. Those tests now read the committed index: NEWEST_ID is qualify.latest_certified, the COMMITTED round trip uses the index's last two ids, the paragraph must sit directly before the subsection's next heading, the synthetic release is one minor version above the newest certified, and the older-version refusal names the live newest version. PE641 and PE650 stay where a test reproduces the hand-built #188 and #190. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Commit the receiver's sealed qualification of pe6.7.5_uk2.130.1_policyengine-uk-data-1.58.1 from replay-state request bc2090402dea079fad906cada6ca93378c7421b7c598ab800787820074b8e6aa (receipt c390b7db49fd2cfde3ac0de048bfbdba12dca08a332a11ed808f4828e90b6070, receipt commit 1291853): its results and release diff and requirements freeze byte for byte, the variant record, the CI matrix entry, the tests that name the newest certified variant and a RELEASE_REPLAY.md paragraph from the receipt's counts. Replay-Request: bc2090402dea079fad906cada6ca93378c7421b7c598ab800787820074b8e6aa
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
The discovery floor cases were written when 6.5.0 was the newest certified variant. #193 pinned the index they pass to candidates(), but discover() reads the committed index through variants.load_document(), so 25 cases failed on the first automated results PR (#195, 6.7.5): with 6.7.5 certified, the floor moved and 6.5.1 and 6.5.2 were no longer discoverable. An autouse fixture now hands discover() the same floor index; a test that needs another index still patches it itself. Checked against #195's index (6.7.5 appended): the old file fails 25 tests, this one passes all 127. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
Author
|
Closing as planned: this was evidence only. The changes it combined have landed separately. #199 and #196 merged, #201 merged as |
This branch was successfully deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Evidence only; it will be closed once CI has run and is never merged.
This branch is #196's head, plus:
5c10b3f8, the PR stacked on Read CI's replay-engine matrix from the variant index; drop the results bot's workflows permission #196);ci.ymltakes #196's version, which reads the matrix from the index. CI here shows whether the whole suite passes once a variant newer than 6.5.0 is certified, which local runs can't check during the disk pause.🤖 Generated with Claude Code