aossie-org/pictopy/pre-pr-check
Claude Code and agent skill by aossie-org/pictopy
- community
- ext:[{"type":"hook","description":"Pre/post lifecycle hooks"}]
Checking that the SKILL.md file is still in the repository.
Run every gate CI runs, before opening a PR. Use before pushing, before opening a pull request, or when asked to verify a change is ready.
Install
npx vskill@latest install aossie-org/pictopy/pre-pr-checkSemi-extensibleMentions customization, but needs skill-specific knowledge. Read the extensible skills standard
SKILL.md
Pre-PR check
Run every check CI runs, in the same order, and report each one. Do not stop at the first failure — run everything, then summarise, so the contributor sees the full picture in one pass.
Two things here are deliberately not a mirror of CI. cargo test is extra: no PR workflow runs it, but CONTRIBUTING.md asks for it on Rust changes. Everything else below is a gate CI will fail you on, including the two PyInstaller builds, which catch imports that work in development and break once frozen.
agent-kit/references/ci-gates.md documents which workflow job each gate comes from. If you change this list, change that file too.
1. Scope
Find what changed:
Skip gates for areas the diff does not touch, and say which ones you skipped and why. A docs-only change does not need cargo test.
2. Markdown
Runs on every .md in the repository regardless of what changed.
The version matches the action CI uses. Leave it pinned, or you will be linting against a different release than the one that decides your build.
Errors from paths inside local virtualenvs (.env/, .docs-env/, .sync-env/, site-packages/) are pre-existing noise, not caused by the change. Ignore them and say so.
3. Agent hook tests
CI runs this on every PR, so run it on every PR too, whatever the diff touches. It needs no dependencies and finishes in under a second.
4. Frontend
Every command below runs from the repository root. Each cd is wrapped in a subshell so it does not leak into the next command.
lint:check runs with --max-warnings 0. The most common failure by far is a leftover TODO or FIXME comment, which no-warning-comments treats as an error. If you see one, remove it — do not disable the rule.
If format:check fails, fix it with cd frontend && npm run format:fix rather than editing by hand.
5. Python
Covers both backend/ and sync-microservice/.
If the black hook reformats files, that is a pass with changes — re-stage them. Never resolve a formatting disagreement by running ruff format; Ruff is configured at 300 columns here and would reflow the entire codebase.
Then the two frozen builds, which CI runs before the tests and which fail more often than the tests do. A dependency that is imported dynamically passes pytest and dies here.
Skip these only for a change that touches no Python at all, and say that you skipped them.
6. Rust
cargo fmt is a CI gate. cargo test is not run by any PR workflow, but CONTRIBUTING.md asks for it, so run it on any src-tauri/ change.
If the format check fails, run cargo fmt without --check to fix it.
7. Repository consistency
- If
git difftouches any ofpackage.json,frontend/package.json, or
frontend/src-tauri/Cargo.toml, confirm all three versions still match. They must only change via npm run version:bump -- X.Y.Z.
- If any file under
agent-kit/skills/changed, confirm the matching stubs exist at
.claude/skills/<name>/SKILL.md and .agents/skills/<name>/SKILL.md, and that the description in each stub matches the playbook's. This is the one place the stub layout can silently rot.
8. Pull request readiness
- The PR body must reference its issue as
#<number>—linked-issue.ymlcopies labels
from the linked issue onto the PR.
- Confirm the issue was reviewed and labelled by a maintainer. PictoPy asks contributors
not to open PRs against unlabelled issues.
- PRs target
dev.
Report
Finish with a table: gate, status, and for failures the exact command to reproduce. State plainly whether the change is ready to push. If something failed, do not describe it as ready.
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingagent-kit/skills/pre-pr-check/SKILL.md
Popularity
- Repository stars
- 304
- Repository forks
- 723
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Extension points
hookPre/post lifecycle hooks
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/aossie-org/pictopy/pre-pr-check)