[only21mil]

Problem

The Thread Transfer Report workflow runs after every pull request CI run. To pick a baseline, resolve() calls github.paginate on listWorkflowRuns for every successful push CI run on the base branch, then keeps the first 20. Main has 2,500 such runs, so each report that gets this far pages up to the API's 1,000-result cap. That took 11 requests in the live check below, to use 20 runs.

Those requests come out of the shared GITHUB_TOKEN quota that .github/actions/ci-changes/action.yml already works to save ("every job that runs this shares the repository's 1,000 requests/hour GITHUB_TOKEN quota"). The workflow ran 1,413 times in the 24 hours before this PR.

Change

Replace the paginated listing with two single-page requests of 20 runs each, made in parallel:

• runs whose head_sha is the PR's base SHA, which the code already puts first;
• the newest runs, minus any at the base SHA.

The ordered list is the same as before: exact-base runs first, then the newest other runs, cut to 20. Any run in the old top 20 that isn't at the base SHA is among the 20 newest runs, so it is still found. One case improves: an exact-base run older than the 1,000-result cap was missed before and is found now.

The comment, artifact checks and outputs don't change.

Scope and approval

There is no issue for this. It is a small, self-contained change to one CI script, with no product code and no user-facing behavior. Output stays the same and API use drops. The ci-changes action already avoids REST calls for the same quota reason.

Verification

Focused test:

node --test .github/scripts/thread-transfer-report.test.cjs

The existing "resolves a fallback PR with a redacted head repo and exact main baseline" test now serves 1,000 main runs from a one-page listWorkflowRuns fake, with the exact-base run at position 900. Its paginate fake no longer handles listWorkflowRuns. On main the test fails with unexpected pagination call at thread-transfer-report.cjs:304. On this branch all 7 tests pass and the baseline is the exact-base run. The test uses no timers and checks outputs, not call arguments.

The updated test fails on main and passes on this branch:

Live check: I ran resolve() from main and from this branch against the real, read-only GitHub API with @octokit/rest 22.0.1 (the lockfile version) for the same five recent pull request CI runs. A request hook counted every GET and rejected any other method. For the four runs that target main, main made 14 requests (11 to listWorkflowRuns) and this branch made 5 (2 to listWorkflowRuns). Outputs were identical, including baseline_run_id and baseline_matches_base=true. The fifth, a fork PR with no open match, returns early after 1 request in both.

Request counts and outputs for the same five CI runs, main ("base") and this branch ("candidate"):

vp lint --report-unused-disable-directives and vp fmt --check on the two changed files: both exit 0 with no findings.

Not checked:

• A real Thread Transfer Report run from this branch. The workflow loads its script from the default branch, so that needs a merge. The live check above runs the same resolve() against the same API.
• The upsertCommentForCurrentHead path. This PR doesn't touch it.

Prior PRs: no open or closed PR or issue makes this change. #22232 by Theo Browne changed other parts of this file. Open #21514 by maria trims a neighboring comment and does not touch this change.

Production fix, tests, PR description and images written with Claude Opus 5.5 in T3 Code Nightly.

🤖 Generated with Claude Opus 5.5 in T3 Code Nightly

[macroscopeapp]

Approvability

Verdict: Approved at 97c2d16

Macroscope's review found this PR approvable — This is a narrowly scoped CI-only optimization that replaces exhaustive workflow-run pagination with two bounded parallel queries and updates coverage for the fallback baseline case. The changed code is confined to the CI publisher and its test harness, with no product-facing behavior or sensitive configuration changes.

^{You can add or adjust custom eligibility rules. Learn more.}

[coderabbitai]

Important

Review skipped

Review was skipped as selected files did not have any reviewable changes.

⚙️ Run configuration

• Source mode: Auto
• Repository: pingdotgg/t3code
• Branch: main
• File: .coderabbit.config.ts
• Commit: 5127302
• Layers: Repository TS config
• Review profile: CHILL
• Plan: Advanced
• Run ID: 76c6494f-a7a7-49d0-91ea-a8feb76ce2a9

📥 Commits

Reviewing files that changed from the base of the PR and between 97c2d16 and fa2557f.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

• Source mode: Auto
• Repository: pingdotgg/t3code
• Branch: main
• File: .coderabbit.config.ts
• Commit: f054d1b
• Layers: Repository TS config
• Review profile: CHILL
• Plan: Advanced
• Run ID: 11a96031-337c-42fb-8419-239e528d8186

📥 Commits

Reviewing files that changed from the base of the PR and between f054d1b and 97c2d16.

📒 Files selected for processing: 2

• .github/scripts/thread-transfer-report.cjs
• .github/scripts/thread-transfer-report.test.cjs

Included review availability: This review used your included allowance. Your plan provides up to 10 included reviews per hour; 9 remain after this review.

📝 Walkthrough

resolve fetches successful workflow runs in parallel: runs for the exact PR base SHA and recent runs on the base branch. It prioritizes exact matches, excludes those matches from the recent-run candidates, and checks artifacts for up to 20 candidates.

Changes: Baseline run selection

Layer / File(s)	Summary
Workflow run queries and candidate ordering .github/scripts/thread-transfer-report.cjs, .github/scripts/thread-transfer-report.test.cjs	resolve makes two single-page queries capped at 20 runs each. It places exact-base runs before recent runs with a different SHA and limits artifact checks to 20 candidates. The test mocks 1,000 runs, applies the SHA filter and page limit, and expects baseline run ID 100.

Priority: ⬇️ Low

Estimated code review effort: 2 (Simple) | ~10 minutes

Change: Refactor

Suggested reviewers: t3dotgg

Merge Risk: ⚪ Minimal · up to 97c2d

The baseline-selection change is mergeable after normal checks; no material regression was identified.

Pre-merge checks: 4

Check name	Status	Explanation
Title check		The title is concise, specific, and accurately describes the main performance change: stopping pagination across all main CI runs.
Description check		The description includes all required sections and provides clear problem context, implementation details, scope rationale, focused verification results, limitations, and agent attribution.
Linked Issues check		Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check		Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches: 💡 1

🧪 Generate unit tests (beta)

• Create a new PR

🛠️ Fix failing CI checks: 💡

• Commit to this branch
• Create a new PR

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}