[vjeux]

User impact

Pull requests cut before #8043 that change a component fail pr-interaction ("playwright.interaction.config.ts does not exist"), and test with it: #8229 (approved), #8055 and #8056 so far. The routing's own rule is that such a head passes with the "Interaction proofs predate this branch" notice.

Problem and solution fit

• Why does it fail? The proofs run on the head, which has no config. The decide step skips such a head only when the pull request's own diff leaves the config alone (one that deletes it still fails), asking pr-scope-touches.sh, which diffs from the merge base to HEAD.
• Why did the helper see the config change? interaction-scope checked out GitHub's merge commit, not the head. On feat(Popover): presentation — a BottomSheet on a phone when asked, as iOS presents a popover #8229 the event's base was 17fead75b9, before test(interaction): real touch input for browser interaction proofs #8043, while the merge commit sits on main, so the diff counted what main gained since, the config included, as the pull request's change.
• Why does that matter? Every such branch is blocked, and any branch whose reported base lags main has main's newer changes counted as its own, which can also put out-of-scope pull requests on 4-core runners.

Checking out the head, as the helper expects and pr-a11y does, fixes it.

Expected behavior and authority

The decide step's contract, in its comment and in .github/scripts/ci-interaction.test.mjs: a head cut before the proofs passes with a notice; one that deletes their config still runs them. spec:AST-030 FR9: routing tests are mutation-sensitive; this adds the missing near miss, a base that lags main.

Smallest restoration

One ref: ${{ github.event.pull_request.head.sha }} on interaction-scope's checkout, with a comment. The decide script, helper and pr-interaction are unchanged.

Evidence

• Reproduction before the change: on main's workflow the new test gives applies=true for a head cut before the proofs with the base it was cut from, as in the feat(Popover): presentation — a BottomSheet on a phone when asked, as iOS presents a popover #8229 run.
• Result after the change: applies=false with the notice, with that base and the current one.
• Representative unchanged path: heads that carry the proofs or delete their config still give applies=true; existing cases are unchanged.
• Regression test or other mutation-sensitive proof: the new test builds the history in a bare origin (a base, main gaining the config, three heads and their merge commits), clones what the job's checkout names, and runs the step with the repository's helper. Removing the ref fails it.

Scope

• One defect is restored.
• Separable contradictory or unsettled tagalongs were removed or split.
• No new public API, visual representation, default, or interaction model is hidden under the bug-fix label.
• Public text and artifacts contain no internal Meta context.

Testing

• pnpm exec vitest run .github/scripts/ci-interaction.test.mjs .github/scripts/ci-duration-guard.test.mjs: 33 passed. The new case failed first on main's workflow (applies=true).
• pnpm exec vitest run .github/scripts: 73 files, 1,335 tests passed.
• pnpm exec prettier --check .github/workflows/ci.yml .github/scripts/ci-interaction.test.mjs: clean.
• Remedy for pull requests already red (feat(Popover): presentation — a BottomSheet on a phone when asked, as iOS presents a popover #8229, chore(core): a compositor animator for touch presentations #8055, chore(core): the animator plans Reduce Motion in its own module, shares a track per element and starts at the commit #8056): once this is on main, close and reopen each. The pull_request trigger lists no types, so the default reopened runs CI again on a new merge commit carrying this workflow, with the head unchanged; a re-run would reuse the old merge commit.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
astryx	Ready	Preview	Oct 25, 2026 10:28pm UTC

[astracat-bot]

AI review status for this pull request.

Review status	Updated
🟢 Reviewed (for maintainers only)	Oct 25, 2026, 10:42 PM UTC

[github-actions]

PR Analysis Report

📚 Storybook Preview

View Storybook for this PR

🧪 Sandbox Preview

View Sandbox for this PR

No new or modified components detected.

Bundle Size Summary

No component packages changed.

Accessibility Audit

Status: No accessibility violations detected.

Visual Regression

48 of 502 shot(s) changed.

A change here is a question, not a failure: check whether the after is the
picture you intended. Record that review on the PR. Baseline maintenance is an
explicit dispatch of CI; this report never rewrites the baseline or adds a release gate.

component	story	theme	mode	pixels
Selector	Indicator Space Bottom Sheet Narrow End RTL	neutral	light	753,933
Selector	Indicator Space Bottom Sheet Wide Start	neutral	light	740,072
RichTextEditor	Markdown parity: side by side	probe	dark	556,900
RichTextEditor	Markdown parity: side by side	probe	light	538,816
Chat	Full AI Chat	probe	light	230,823
Chat	Full AI Chat	probe	dark	229,805
AppShell	With Mobile Nav	probe	light	217,010
AppShell	With Mobile Nav	probe	dark	216,960
Collapsible	Dividers	probe	dark	66,445
Collapsible	Dividers	probe	light	66,445
AppShell	Playground	neutral	dark	65,198
AppShell	Playground	neutral	light	63,788
Chat	Default	neutral	dark	36,304
Chat	Default	neutral	light	35,767
Chat	Empty State	probe	light	23,531
Chat	Empty State	probe	dark	22,622
ChatAutoScroll	Scroll Behavior Comparison	neutral	dark	15,378
ChatAutoScroll	Scroll Behavior Comparison	neutral	light	14,107
Chat	With Attachments	probe	light	13,406
Chat	With Attachments	probe	dark	11,232

and 28 more.

---

_{Generated by PR Enrichment workflow | Storybook | Sandbox | View full report}

[vjeux]

Superseded by #8812, which makes interaction-scope check out the pull request head (and native-swift-scope the head or the queued merge), with routing tests that include a head cut before the proofs. Closing this one; the branch stays.