ci: single-leg coverage and cached utoo package store

[killagu]

Motivation

Cut CI wall time on the 3-OS × 2-Node test matrix without adding matrix jobs (sharding is intentionally out of scope — it multiplies jobs and queues the runner pool at this repo's PR volume). Two independent, low-risk levers.

Task A — collect coverage on one matrix leg only

ut run ci (= ut run test -- --coverage) ran v8 coverage on all 6 legs, but codecov only needs one upload. Coverage adds ~10–30% per-leg overhead.

• Add a coverage matrix axis (default false), enabled on a single ubuntu/node-24 leg.
• Split "Run tests" into two guarded steps — ut run ci (with --coverage) on the coverage leg, ut run test on the other 5. Both reach the same pretest prep (clean-dist + workspace pretest); only --coverage differs. (ut run ci only reaches clean-dist via its nested ut run test, so dropping ci for test keeps the prep identical.)
• Gate the codecov upload on matrix.coverage (was matrix.os != 'windows-latest'), so exactly one leg uploads. Applied to test, test-egg-bin, and test-egg-scripts.

Note

exclude is required, not just include. GitHub's include cannot overwrite an existing matrix dimension. With coverage: [false] in the base, include: { coverage: true } would create a separate ubuntu/24 leg → 7 legs, ubuntu/24 run twice. The added exclude of the base ubuntu/24/coverage:false combo makes the coverage leg replace it. Leg counts stay 6 / 2 / 2; the done gate is unaffected.

Task B — cache the utoo package download store

Verified where ut install --from pnpm actually reads from before caching:

• It clones from a content-addressable store at $HOME/.cache/nm/<pkg>/<version>/ (default), redirectable via the UTOO_CACHE_DIR env var (both confirmed empirically — a fresh uncached install populated exactly those dirs; warm installs report cloned, not downloaded). Not a no-op.
• Pin the store to ${{ runner.temp }}/utoo-store (deterministic, identical expression in env and the cache path, outside the checkout) and wrap ut install with actions/cache in every installing job (typecheck, test, test-egg-bin, test-egg-scripts).

Note

Cache key deviates from a lockfile hash. This repo commits no lockfile — dependency versions live in pnpm-workspace.yaml's catalog + the package.json files (utoo generates an uncommitted package-lock.json at install time). hashFiles('**/pnpm-lock.yaml') would be empty → a constant key that never re-saves the growing store. The key hashes pnpm-workspace.yaml + **/package.json instead, with restore-keys: pkgstore-<os>- falling back to the latest per-OS store on a key miss.

Per repo convention, every action is pinned by commit SHA (actions/cache@0057852… = v4.3.0).

Test evidence

• actionlint passes on the workflow.
• Matrix expansion simulated against GitHub's documented include/exclude algorithm: 6 / 2 / 2 legs (unchanged), exactly 1 coverage leg each (ubuntu/24).

Before/after timings — to fill in from this PR's first runs

These require real CI runs and will be added once the workflow executes here:

• 5 non-coverage test legs: duration vs. next baseline (Task A).
• warm-cache ut install step duration vs. cold (Task B); cache hit-rate visible in the Actions cache UI.

🤖 Generated with Claude Code

Summary by CodeRabbit

• Chores
  • Improved CI reliability by refining how dependencies are cached during validation runs.
  • Streamlined test execution with an explicit coverage vs. non-coverage flow, ensuring coverage runs only for the intended OS/node combinations.
  • Reduced redundant coverage reporting by uploading coverage results only when running in coverage-enabled mode.

[gemini-code-assist]

Note

Gemini is unable to generate a review for this pull request due to the file types involved not being currently supported.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: b1af916a-70ee-4f14-8b13-6cebfedde43c

📥 Commits

Reviewing files that changed from the base of the PR and between a39c3be and 46e83bb.

📒 Files selected for processing (1)

• .github/workflows/ci.yml

🚧 Files skipped from review as they are similar to previous changes (1)

• .github/workflows/ci.yml

---

📝 Walkthrough

Walkthrough

The CI workflow adds utoo package store caching with UTOO_CACHE_DIR in typecheck and test jobs, and introduces a coverage matrix dimension for test, test-egg-bin, and test-egg-scripts. Those jobs now branch between ut run ci and ut run test, with Codecov uploads gated by matrix.coverage.

Changes

CI Cache and Coverage Matrix

Layer / File(s)	Summary
utoo cache in typecheck job .github/workflows/ci.yml	Adds actions/cache for the utoo package store and sets UTOO_CACHE_DIR during dependency install in the typecheck job.
coverage matrix + steps for test job .github/workflows/ci.yml	Adds a coverage matrix branch for the test job, caches the utoo package store, splits ut run ci and ut run test, and gates Codecov upload on matrix.coverage.
coverage matrix + steps for test-egg-bin job .github/workflows/ci.yml	Applies the same coverage matrix, utoo cache, execution split, and Codecov gate changes to test-egg-bin.
coverage matrix + steps for test-egg-scripts job .github/workflows/ci.yml	Applies the same coverage matrix, utoo cache, execution split, and Codecov gate changes to test-egg-scripts.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• eggjs/egg#5845: Also changes .github/workflows/ci.yml test-job matrix/execution shape and CI command selection.

Poem

🐇 I hop through cache with speedy feet,
The coverage path makes CI neat.
One branch runs tests, one shines with trace,
Codecov lands in the proper place.
A rabbit approves this matrix dance 🌙

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly summarizes the CI coverage split and utoo package store caching changes.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch feat/ci-single-leg-coverage-cache

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[cloudflare-workers-and-pages]

Deploying egg with    Cloudflare Pages

Latest commit:	46e83bb
Status:	✅  Deploy successful!
Preview URL:	https://0ac94a39.egg-cci.pages.dev
Branch Preview URL:	https://feat-ci-single-leg-coverage.egg-cci.pages.dev

View logs

[cloudflare-workers-and-pages]

Deploying egg-v3 with    Cloudflare Pages

Latest commit:	46e83bb
Status:	✅  Deploy successful!
Preview URL:	https://84664ac8.egg-v3.pages.dev
Branch Preview URL:	https://feat-ci-single-leg-coverage.egg-v3.pages.dev

View logs

[socket-security]

Dependency limit exceeded — report not shown.

This pull request scan exceeded the 10,000-dependency limit applied to this scan, so the results are incomplete and may be inaccurate. To avoid reporting false positives, Socket has not posted a report.

Upgrade your plan to raise the dependency limit and get complete reports, or view the partial scan in the dashboard.

Socket is always free for open source. If this is a non-commercial open source project, contact us to request a free Team account.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.99%. Comparing base (22d5585) to head (46e83bb).

Additional details and impacted files

@@            Coverage Diff             @@
##             next    #6006      +/-   ##
==========================================
- Coverage   82.85%   81.99%   -0.87%     
==========================================
  Files         676      676              
  Lines       20522    20522              
  Branches     4060     4060              
==========================================
- Hits        17004    16826     -178     
- Misses       3040     3189     +149     
- Partials      478      507      +29     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.