Use when committing code - ensures atomic, descriptive commits that leave the codebase in a merge-ready state at every point
Ensures every commit is atomic, descriptive, and leaves code in a working state. Automatically invoked when committing code to enforce proper message format, test verification, and single-logical-change principles.
/plugin marketplace add troykelly/claude-skills/plugin install issue-driven-development@troykelly-skillsThis skill is limited to using the following tools:
Every commit is atomic, descriptive, and leaves code in a working state.
Core principle: Anyone should be able to checkout any commit and have working code.
Announce at use: "I'm committing with a descriptive message following clean-commits standards."
[type](scope): Short description (max 72 chars)
[Optional body - what and why, not how]
[Optional footer - issue references, breaking changes]
Refs: #[ISSUE_NUMBER]
| Type | Use For |
|---|---|
feat | New feature |
fix | Bug fix |
docs | Documentation only |
style | Formatting, no code change |
refactor | Code restructuring |
test | Adding/fixing tests |
chore | Maintenance, dependencies |
feat(auth): Add user registration endpoint
Implement POST /api/users/register with email validation,
password hashing, and duplicate detection.
- Validates email format and uniqueness
- Hashes password with bcrypt
- Returns user object without password
Refs: #123
fix(auth): Prevent redirect loop on expired session
Session expiry was triggering redirect to login, which
checked session, found expired, and redirected again.
Now clears session cookie before redirecting.
Refs: #456
test(auth): Add integration tests for registration
Cover success case, duplicate email, invalid format,
and weak password scenarios.
Refs: #123
| Atomic | Not Atomic |
|---|---|
| One logical change | Multiple unrelated changes |
| Passes all tests | Breaks tests |
| Complete feature slice | Half-implemented feature |
| Can be reverted cleanly | Reverts would break things |
If you have multiple changes, commit them separately:
# Stage specific files
git add src/auth/register.ts
git add src/auth/register.test.ts
git commit -m "feat(auth): Add registration endpoint"
# Stage next logical unit
git add src/auth/login.ts
git add src/auth/login.test.ts
git commit -m "feat(auth): Add login endpoint"
Every commit must leave the codebase in a state where:
Before committing:
# Run tests
pnpm test
# Check build
pnpm build
# Check types
pnpm typecheck
# Check lint
pnpm lint
If any fail, fix before committing.
| Too Infrequent | Just Right |
|---|---|
| "Implement entire feature" | "Add user model" |
| "Fix all bugs" | "Fix session expiry redirect" |
| "Update everything" | "Update auth dependencies" |
Smaller commits are:
# Review what changed
git diff
# Stage specific files
git add [specific files]
# Or stage interactively
git add -p
# See what will be committed
git diff --staged
# Short message (if simple)
git commit -m "fix(auth): Handle null user in session check"
# Long message (if complex)
git commit
# Opens editor for full message
# Check commit looks right
git show --stat
# Verify tests still pass
pnpm test
When to include a body:
refactor(api): Extract validation middleware
Validation logic was duplicated across 12 endpoints.
Extracted to reusable middleware that can be composed.
Alternative considered: validation library.
Rejected because our rules are domain-specific.
fix(data): Use optimistic locking for updates
Race condition was causing lost updates when two users
edited the same record simultaneously.
BREAKING CHANGE: Update operations now require
version field in request body.
Always reference the issue:
# In commit message
Refs: #123
# Or if commit closes the issue
Closes: #123
# Amend last commit (before push only!)
git add forgotten-file.ts
git commit --amend
If a commit was wrong:
# Create a new commit that undoes the change
git revert [commit-sha]
# DON'T rewrite history on shared branches
# DON'T force push to fix mistakes
Before each commit:
This skill is called by:
issue-driven-development - Throughout developmentpr-creation - Before creating PRThis skill enforces:
This skill should be used when the user asks to "create an agent", "add an agent", "write a subagent", "agent frontmatter", "when to use description", "agent examples", "agent tools", "agent colors", "autonomous agent", or needs guidance on agent structure, system prompts, triggering conditions, or agent development best practices for Claude Code plugins.
This skill should be used when the user asks to "create a slash command", "add a command", "write a custom command", "define command arguments", "use command frontmatter", "organize commands", "create command with file references", "interactive command", "use AskUserQuestion in command", or needs guidance on slash command structure, YAML frontmatter fields, dynamic arguments, bash execution in commands, user interaction patterns, or command development best practices for Claude Code.
This skill should be used when the user asks to "create a hook", "add a PreToolUse/PostToolUse/Stop hook", "validate tool use", "implement prompt-based hooks", "use ${CLAUDE_PLUGIN_ROOT}", "set up event-driven automation", "block dangerous commands", or mentions hook events (PreToolUse, PostToolUse, Stop, SubagentStop, SessionStart, SessionEnd, UserPromptSubmit, PreCompact, Notification). Provides comprehensive guidance for creating and implementing Claude Code plugin hooks with focus on advanced prompt-based hooks API.