[m4ttheweric]

rt#475 missed both time targets (full 3m56s, later 3m27s on main; typescript-only 2m07s). the step timings say why: static carried the 60s Go test step on every PR's critical path, and the committed shard timings were measured on a laptop where shard 2 ran 163s but 95s on the runner. this fixes both and takes the two hygiene items the review deferred.

what changed

• RT-323: rt-ui vet and tests moves out of static into its own ubuntu go job; checks requires it. a manual Timings workflow (gh workflow run timings.yml) runs bun run test:timings on macos-latest and uploads test-timings.json; it also runs once when the workflow file itself changes, which is how this PR's committed timings were measured. AGENTS.md says to refresh from the workflow, not a laptop.
• RT-324: the four source scanners the no-* glob missed are renamed into it (no-spawn-without-env, no-legacy-state-sources, no-unlisted-command-call-sites, and no-hand-built-repo-paths split out of rt-paths.test.ts); alwaysRun() walks every unit directory for no-*.test.ts and the explicit list is gone.
• RT-325: test:watch and test:all delegate to bun run test; the parity test asserts no other script restates the directory list.

verification

measurements land in a comment once the Timings artifact is committed and the run is green.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

Next included review available in 4 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used the included review currently available. Your 86 included PR review attempts over the past 7 days set your current allowance at 1 review per hour.

Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Advanced

Run ID: 92f1a62d-8465-4406-bc4f-2d0c165d76c3

📥 Commits

Reviewing files that changed from the base of the PR and between f467648 and dc7d173.

📒 Files selected for processing (14)

• .github/workflows/checks.yml
• .github/workflows/timings.yml
• AGENTS.md
• lib/__tests__/flavor.test.ts
• lib/__tests__/no-hand-built-repo-paths.test.ts
• lib/__tests__/no-spawn-without-env.test.ts
• lib/__tests__/rt-paths.test.ts
• lib/rt-paths.ts
• lib/state/__tests__/no-legacy-state-sources.test.ts
• package.json
• packages/rt-client/test/no-unlisted-command-call-sites.test.ts
• scripts/ci/__tests__/test-scope.test.ts
• scripts/ci/test-scope.ts
• test-timings.json

---

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

[m4ttheweric]

measurements after merge (Checks workflow only)

run	mode	wall	shards	notes
main 86f2936 (this merge)	full	3m21s	143s / 158s / 142s	runner-measured timings balance within 16s; static 44s, go 56s off the critical path
rt#481 typescript-only probe (hydrate.ts newline)	changed	1m26s	30s / 29s / 38s	11 always-run guards in 2.8s on shard 1

before this PR: full 3m56s (rt#475) and 3m27s (main f467648), typescript-only 2m07s. the typescript-only target (under 2 min) is met with 34s to spare. the full run sits at about 3 min; the remaining 20s over is shard setup (checkout, setup-bun, brew, bun install, about 22s) plus runner pickup, and a fourth shard would push a main push past the org's five concurrent macOS jobs whenever the pty gate also runs.