Workflow
Skills I use with Claude Code across my projects. Architecture, code review, testing, security, deployment, and more.
npx -y skills add pvnarp/agent-skills --skill workflowAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Enforces git branching strategy, PR process, and merge rules. Covers branch naming, commit conventions, PR creation, CI requirements, and the merge-to-main checklist. Use when creating branches, making commits, opening PRs, or merging code.
SKILL.md
2.8 KB, 663 tokens by cl100k_base, as published. Nobody here has run it
Git Workflow
Main is always deployable. No exceptions.
Branching Strategy (GitHub Flow)
main (protected - always deployable)
├── feature/user-auth
├── feature/search-api
├── fix/login-timeout
└── chore/upgrade-deps
Branch Naming
| Prefix | Use |
|---|---|
feature/ | New functionality |
fix/ | Bug fixes |
refactor/ | Code restructuring with no behavior change |
chore/ | Build config, dependencies, CI, docs |
Rules:
- Lowercase, hyphens only. No spaces, no underscores.
- Short and descriptive:
feature/user-search, notfeature/implement-the-user-search-functionality-with-filters. - One concern per branch. Don't mix a feature with a refactor.
Creating a Branch
git checkout main
git pull origin main
git checkout -b feature/your-feature
Commit Conventions
Format: type: short description
feat: add user search endpoint
fix: correct pagination off-by-one
refactor: extract auth middleware into separate module
chore: upgrade dependencies to latest
test: add unit tests for rate limiting
Types: feat, fix, refactor, chore, test, docs
Rules:
- Present tense, imperative mood ("add" not "added" or "adds").
- First line under 72 characters.
- Commit often. Small, atomic commits that each do one thing.
- Never commit:
.env, credentials, API keys, large binaries.
Pull Request Process
1. Before Opening PR
- All commits follow conventions
- Code builds without errors
- All tests pass
- No new warnings introduced
- Self-review: read your own diff
2. Open PR
gh pr create --title "feat: add user search" --body "$(cat <<'EOF'
## Summary
- Brief description of what changed and why
## Test plan
- [ ] Unit tests pass
- [ ] Manual verification of key flows
EOF
)"
3. CI Must Pass
Automated checks run on every PR. All must pass before merge.
4. Code Review
Run /review on the PR changes. Address all CRITICAL and WARNING items.
5. Merge to Main
gh pr merge --squash
Squash merge keeps main history clean. The PR title becomes the merge commit message.
6. Clean Up
git checkout main
git pull origin main
git branch -d feature/your-feature
What Goes Directly to Main (Never)
Nothing. Even a one-line fix gets a branch and PR. The CI gate and review step catch things humans miss.
Hotfix Process
Same as above but with urgency:
git checkout main && git pull
git checkout -b fix/critical-issue
# fix, commit, push, PR, CI passes, merge
Gives 2 of the 12 instructions most pr commit review skills give in 663 tokens
Counted across 888 of the 1,342 authors here whose files we hold, read 2026-08-07
- use conventional commits formatin 127 of 888, across 115 files
- keep subject line under 72 charactersin 62 of 888, across 48 files
- delete branches after mergehere, and in 51 of 888, across 38 files
- use imperative mood in subject linein 51 of 888, across 42 files
- use imperative mood in commit messagesin 44 of 888
- verify directory is ignored before creating worktreein 43 of 888, across 12 files
- generate a conventional commit messagein 43 of 888
- add unignored worktree directories to gitignorein 42 of 888, across 10 files
- make atomic commitshere, and in 39 of 888, across 27 files
- run tests before committingin 36 of 888, across 25 files
- verify clean test baselinein 35 of 888, across 9 files
- split unrelated changes into separate commitsin 35 of 888, across 30 files
Said here and by no other author read
- write present tense imperative commit messages
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.