agentleFS
kaeawckaeawc/auto-mobile/skills/node-prepush/SKILL.md

node-prepush

Skill56 starsChanged 34 days agoApache-2.0View on GitHub

Run the Node pull-request gates before pushing TypeScript changes, and triage verified flakes without speculative test changes.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this workflow before pushing a Node or TypeScript change.

    line 8
  2. Very specific:

    Run bash scripts/prepush-node.sh.

    line 10
  3. Very specific:

    Treat a non-zero bun run lint exit as a failure.

    line 13

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Node Pre-push

Use this workflow before pushing a Node or TypeScript change.

  1. Run bash scripts/prepush-node.sh. During a short local feedback loop, use bash scripts/prepush-node.sh --changed; it still runs format, typecheck, and lint in full because their baseline and boundary gates are global.
  2. Treat a non-zero bun run lint exit as a failure. Error-level oxlint rules such as eqeqeq are direct gates; the warning ratchet cannot excuse them.
  3. The first gate fetches origin/main and requires the current HEAD to contain it. Rebase or merge main, then re-run the gate before requesting a merge.
  4. For a suspected Node CI flake, inspect the exact failed job log and rerun only after evidence supports it. This week's verified loaded-runner 100ms timing overages use the timing gate's isolated recheck; do not alter tests when that recheck is clean. Treat a matching cross-platform failure as shared state or stale base until proven otherwise.

What the gate reproduces, and what stays CI-only

prepush-node.sh runs: stale-base guard, format:check, typecheck, lint (oxlint, ratchet, check-boundaries.sh), typecheck:test, tool definitions drift, the static Fast Validation subset (yaml, runtime-pins, stdlib-first, shared-process-allowlist, docs-assets, env-var-docs, lfs-pointers, lychee-offline, and the other tool-free checks named in FAST_VALIDATION_CHECKS), repository lint tests, the image smoke, the unit lane, the test/db fast path, and the 100 ms timing budget on Linux.

Repository lint tests run only test/lint/*.integration.test.ts in full mode; the unit files already run in the unit lane. In --changed mode, early affected-test selection supplies branch, worktree, and untracked paths; the whole test/lint/ directory runs only for changes under src/, test/, scripts/, oxlint-plugins/, android/, ios/, docs/, skills/, schemas/, .claude/, .github/, or root .oxlintrc.json, package.json, bun.lock, bunfig.toml, turbo.json, or tsconfig*.json. After that directory runs, affected unit and integration extras omit test/lint/ files already covered. Always-run unit extras still run when repository lint tests skip. A unit cap fallback replaces the changed lane with the full unit lane and skips unit extras; always-run bats and affected safe integration lanes still run afterward. The timing gate measures the unit lane's JUnit reports when source or runtime inputs changed; changed test files alone are still re-run individually. Changed mode without reports uses the timing validator's own affected-test selection. Typecheck and lint stay strictly serial and separate because oxlint 1.88's experimental --type-check and tsgolint diagnostics do not replace the fail-closed typecheck parser, TS2688/@types/bun guard, or separate src/test tsconfig projects.

CI-only by design: the Windows and macOS unit/integration legs (path separators, test/lint/pathSeparatorAssertions.test.ts is the static proxy), the loaded-runner timing budget (a local pass proves only the unloaded case; --timing forces it on macOS), ShellCheck at CI's pinned version (local may be newer: keep # shellcheck disable=SC2329 paired with SC2317), online docs link checks (merge.yml), the clean-room pinned runtime graph install (scripts/ci/verify-pinned-runtime-graph.sh, needs network), benchmarks, CodeQL, and the Android/iOS/desktop builds.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.