agentleFS
hhhyyyy99hhhyyyy99/prot-skills/.agents/skills/self-check/SKILL.md

self-check

Skill7 starsChanged 5 months agoMITView on GitHub

Run local production/PR self-checks before pushing or submitting a PR. Mirrors pre-push checks and the GitHub PR CI checks where practical. Use when asked to "自检", "生产门禁自检", "跑门禁", "pre-push check", "PR checks", "before push", or "before PR".

  • Installs packages

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

17 rules. 0 are common, 11 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 11

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Before submit-pr or any request to push changes

    line 12
  2. Distinctive:

    After committing feature, fix, refactor, test, docs, config, or release-prep changes

    line 13
  3. Distinctive:

    When the user asks to run "自检", "生产门禁", "PR 检测", or "pre-push"

    line 14
  4. Distinctive:

    When checking whether the current branch is ready for GitHub CI

    line 15
  5. Distinctive:

    Check branch and working tree:

    line 62
  6. Distinctive:

    If the working tree is dirty, continue only if the user asked to check the current local state.

    line 74
  7. Very specific:

    Run exactly the commands from .husky/pre-push:

    line 82
  8. Distinctive:

    Do not fix unrelated lint warnings during self-check.

    line 100
  9. Distinctive:

    Run the frontend CI commands that are not already fully covered by pre-push:

    line 108
  10. Distinctive:

    Run the backend CI commands:

    line 126
  11. Distinctive:

    Prefer separate commands so failures are easy to identify.

    line 134
  12. Distinctive:

    If local Linux system dependencies are missing, report the blocker and the CI packages required.

    line 136
  13. Very specific:

    Use the run or verify skill if available, otherwise use the project commands:

    line 146
  14. Distinctive:

    Only do this when the user asked for runtime verification or the change is UI/interaction behavior where tests/builds are not enough.

    line 154
  15. Very specific:

    pnpm install --frozen-lockfile would fail in CI: if lockfile/package changes are present, run or recommend pnpm install --frozen-lockfile …

    line 225
  16. Very specific:

    pnpm app:check fails: run pnpm app:sync, inspect metadata changes, and commit them if expected.

    line 226
  17. Very specific:

    cargo fmt --check fails: run cd src-tauri && cargo fmt, then review and commit formatting changes.

    line 228

“Common” is measured: kept by at least 6 in 10 of the 8 edited versions of this file we could read. “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

Self Check

Runs the repository's local production gate and the PR CI checks that can be reproduced locally. Use this before pushing a branch or opening a PR.

When to Use

  • Before submit-pr or any request to push changes
  • After committing feature, fix, refactor, test, docs, config, or release-prep changes
  • When the user asks to run "自检", "生产门禁", "PR 检测", or "pre-push"
  • When checking whether the current branch is ready for GitHub CI

What PR Runs

GitHub PR checks are defined in .github/workflows/ci.yml and .github/workflows/ai-review.yml.

PR CheckLocal equivalentNotes
Frontend Checks → pnpm app:checkpnpm app:checkVerifies app metadata sync
Frontend Checks → pnpm testpnpm testVitest frontend/script suite
Frontend Checks → pnpm buildpnpm buildRuns app:sync, tsc, and Vite build
Backend Checks → pnpm app:checkAlready covered by pnpm app:checkCI runs it again in backend job
Backend Checks → pnpm buildAlready covered by pnpm buildBuilds frontend assets for Tauri context
Backend Checks → cargo fmt --checkcd src-tauri && cargo fmt --checkRequires Rust toolchain
Backend Checks → cargo clippy --all-targets --all-features -- -D warningsSame command locallyRequires system Tauri/WebKit dependencies on Linux
Backend Checks → cargo test --all-featuresSame command locallyRequires Rust toolchain
ai-reviewDo not run locally by defaultRequires PR context, GitHub token, and AI review secrets

Local pre-push gate is defined in .husky/pre-push:

pnpm test
pnpm lint
pnpm format:check
pnpm test:tokens

Prerequisites

  • Clean working tree is preferred; if dirty, report the changed files before running checks.
  • Dependencies are installed with pnpm install.
  • Rust toolchain is available for backend checks.
  • On Linux, backend clippy/tests may require the same system packages CI installs:
    • libwebkit2gtk-4.1-dev
    • libgtk-3-dev
    • libayatana-appindicator3-dev
    • librsvg2-dev
    • patchelf

Workflow

Execute these phases in order. Report each phase result before moving on if it fails.


Phase 1 — Inspect Current State

  1. Check branch and working tree:

    git status --short --branch
  2. Inspect recent commits when relevant:

    git log --oneline -5
  3. If the working tree is dirty, continue only if the user asked to check the current local state. Otherwise, ask whether to commit/stash first.

Output: branch name, clean/dirty state, and whether checks are being run against committed-only or working-tree changes.


Phase 2 — Run Local Pre-Push Gate

Run exactly the commands from .husky/pre-push:

pnpm test
pnpm lint
pnpm format:check
pnpm test:tokens

A convenient single command is:

pnpm test && pnpm lint && pnpm format:check && pnpm test:tokens

Notes:

  • pnpm lint may report warnings. Warnings are acceptable if it exits successfully and reports 0 errors.
  • Do not fix unrelated lint warnings during self-check. Report them separately if they are relevant.

Output: pass/fail for tests, lint, format, and token consistency.


Phase 3 — Run PR Frontend CI Equivalents

Run the frontend CI commands that are not already fully covered by pre-push:

pnpm app:check
pnpm build

Notes:

  • pnpm build already runs pnpm app:sync, then tsc, then vite build.
  • pnpm test was already covered in Phase 2; do not rerun it unless a previous command changed files.

Output: app metadata check and frontend build result.


Phase 4 — Run PR Backend CI Equivalents

Run the backend CI commands:

cd src-tauri && cargo fmt --check
cd src-tauri && cargo clippy --all-targets --all-features -- -D warnings
cd src-tauri && cargo test --all-features

Prefer separate commands so failures are easy to identify. Avoid changing directories permanently; run commands with an explicit working-directory pattern when possible.

If local Linux system dependencies are missing, report the blocker and the CI packages required. Do not install system packages unless the user explicitly asks.

Output: Rust format, clippy, and test results.


Phase 5 — Optional Runtime Verification

For UI changes, run the app and observe the changed surface before claiming feature-level readiness.

Use the run or verify skill if available, otherwise use the project commands:

pnpm dev
# or
pnpm tauri:dev

Only do this when the user asked for runtime verification or the change is UI/interaction behavior where tests/builds are not enough.

Output: what was launched, what user-visible behavior was observed, and any platform not available locally.


Phase 6 — Report

Report in this format:

Self-check result: PASS | FAIL | BLOCKED

Branch: <branch>
Working tree: clean | dirty (<summary>)

Checks:
- pre-push: PASS | FAIL
  - pnpm test: ...
  - pnpm lint: ...
  - pnpm format:check: ...
  - pnpm test:tokens: ...
- frontend PR CI: PASS | FAIL
  - pnpm app:check: ...
  - pnpm build: ...
- backend PR CI: PASS | FAIL | BLOCKED
  - cargo fmt --check: ...
  - cargo clippy ...: ...
  - cargo test --all-features: ...
- ai-review: not run locally (<reason>)

Notes:
- <warnings, environment gaps, runtime verification notes>

Quick Reference

Full local PR self-check:

pnpm test && pnpm lint && pnpm format:check && pnpm test:tokens
pnpm app:check
pnpm build
(cd src-tauri && cargo fmt --check)
(cd src-tauri && cargo clippy --all-targets --all-features -- -D warnings)
(cd src-tauri && cargo test --all-features)

Pre-push only:

pnpm test && pnpm lint && pnpm format:check && pnpm test:tokens

Frontend PR CI only:

pnpm app:check && pnpm test && pnpm build

Backend PR CI only:

pnpm app:check && pnpm build
(cd src-tauri && cargo fmt --check)
(cd src-tauri && cargo clippy --all-targets --all-features -- -D warnings)
(cd src-tauri && cargo test --all-features)

Error Handling

  • pnpm install --frozen-lockfile would fail in CI: if lockfile/package changes are present, run or recommend pnpm install --frozen-lockfile before PR.
  • pnpm app:check fails: run pnpm app:sync, inspect metadata changes, and commit them if expected.
  • pnpm build fails in TypeScript: fix type errors before rerunning.
  • cargo fmt --check fails: run cd src-tauri && cargo fmt, then review and commit formatting changes.
  • cargo clippy fails: fix the reported warning/error; CI treats all clippy warnings as errors.
  • Linux Tauri dependency missing: report the missing package and the CI install list; ask before installing system packages.
  • AI review not runnable locally: this is expected unless PR number, GitHub token, and AI review secret environment are available.

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.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

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.