Repository navigation
Thread Report: copy the v5 sidecar, so the comment keeps its author counts - #75
Merged
Merged
Conversation
…ounts Every Thread Report this repository has posted said this, directly under the verdict line: No author counts: the v5 manifest was not found beside the head manifest. thread-report.py reads authorship from the v5 manifest sitting beside --head. This workflow copies only trace-manifest.json to /tmp/head.json, leaving trace-manifest.v5beta.json in the app directory, so the sentence ruled in FR-THREAD-20 could never render. Seen on the live comment on #74, 2026-10-09. One line fixes it: copy the sidecar to /tmp/head.v5beta.json. The adopters' path was never broken. ci-thread-report.sh, which the bundle ships and which the cold path installs, has always copied the sidecar (lines 91 to 92). This workflow predates that script and still does the copying by hand, which is how the two diverged. That duplication is the underlying finding and it is not fixed here: replacing these steps with the shipped script would delete the divergence rather than patch it, and it is a bigger change to the workflow the specimens' comments depend on. Two tests, in tests/test_samples.py: one asserts this workflow copies both files, one asserts the shipped script still does. The first fails against main with the workflow reverted and the test kept, which is the control that matters. Suite 224 to 226. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Vzf6Sh5RVMKdFqaXKFF2FY
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
One line, plus two tests so it cannot drift back.
What was wrong
Every Thread Report this repository has ever posted carried this directly under the verdict line:
thread-report.pyreads authorship from the v5 manifest sitting beside--head. The workflow copied onlytrace-manifest.jsonto/tmp/head.jsonand lefttrace-manifest.v5beta.jsonin the app directory, so the authorship sentence ruled in FR-THREAD-20 could never render. Seen on the live comment on #74, 2026-10-09.The fix is the one line you named:
The adopters' path was never broken, and that is the real finding
extensions/specassay-check/scripts/ci-thread-report.sh, which the bundle ships and the cold path installs, has always copied the sidecar:So no adopter has lost the sentence. Only this repository has, because this workflow predates that script and still does the copying by hand. The duplication is what let the two diverge.
I have not fixed that here. Replacing these steps with a call to the shipped script would delete the divergence rather than patch it, and it is a larger change to the workflow the six specimens' comments depend on, on a day the Sites room is shooting from one of them. It is yours to call, and I would rather raise it than quietly widen a one-line fix.
Tests
In
tests/test_samples.py:test_the_demo_workflow_copies_the_v5_sidecar_beside_the_head_manifesttest_the_shipped_ci_copies_the_sidecar_too— the path that was never broken, pinned so it stays that wayControl: with the workflow reverted to
mainand the test kept, the first one fails. That is the direction that matters; reverting both together proves nothing, which is a mistake I made on the first attempt and then redid properly.Verified
OK (154 registry IDs)v0.5.6, 0 warningsOnce this merges I will rebase
demo/close-edit-undoontomainand push, so #74's sticky comment re-renders with the authorship sentence, and report the comment URL again.🤖 Generated with Claude Code
https://claude.ai/code/session_01Vzf6Sh5RVMKdFqaXKFF2FY
Generated by Claude Code