From agentic-doc
This skill should be used when the user asks to "write a commit message", "generate a commit", "help me commit", "create a commit message for", "commit these changes", or mentions commits format. Provides guidance for generating compact, clear commit messages following Conventional Commits specification.
How this skill is triggered — by the user, by Claude, or both
Slash command
/agentic-doc:agd-conventional-commitsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Generate commit messages that are compact, clear, and follow the Conventional Commits specification.
Generate commit messages that are compact, clear, and follow the Conventional Commits specification.
<type>[optional scope]: <description>
[optional body]
[optional footer(s)]
Key rules:
| Type | Usage | Example |
|---|---|---|
feat | New feature | feat: add user authentication |
fix | Bug fix | fix: resolve null pointer in login |
docs | Documentation only | docs: update API endpoints |
style | Formatting, no code change | style: format indentation |
refactor | Code change without fix/feature | refactor: extract utility function |
test | Adding/updating tests | test: add unit tests for auth |
chore | Build, CI, dependencies | chore: bump dependencies |
perf | Performance improvement | perf: optimize query execution |
ci | CI/CD configuration | ci: add GitHub Actions workflow |
build | Build system changes | build: update webpack config |
revert | Revert previous commit | revert: remove broken feature |
Scope indicates the affected module or component:
feat(auth): add OAuth2 support
fix(api): handle timeout errors
docs(readme): add installation steps
Keep scope short - typically one word.
For breaking changes, add ! after type/scope or include BREAKING CHANGE: in footer:
feat!: redesign API endpoints
BREAKING CHANGE: all endpoints now require authentication
feat: add password reset functionality
fix: resolve memory leak in worker process
docs: clarify installation requirements
refactor(search): extract ranking algorithm
test: add edge cases for date parser
chore(deps): bump eslint to 9.0.0
feat(api)!: change response format to JSON
Fixed the bug # Past tense, vague
Added new feature for user login. # Capitalized, period
feat: I added a new login page # First person
fix: fix the thing # Vague, not descriptive
updates to the codebase # No type, vague
feat(authentication-system): add oauth # Scope too long
git diff --stagedUse body for additional context when the change is complex.
Add body when ANY of these conditions are met:
| Condition | Threshold | Example |
|---|---|---|
| Large changes | >100 lines OR >5 files | Refactoring multiple modules |
| Complex logic | Non-obvious changes | Bug fixes, algorithm changes |
| Breaking changes | API changes, config updates | Removing deprecated endpoints |
| Multiple concerns | More than one focus | Fix + refactor in same commit |
Keep description-only when:
When adding body, include:
feat(api): add rate limiting to public endpoints
Implement token bucket algorithm for all public API endpoints.
Limit: 100 requests per minute per IP address.
This protects against abuse while maintaining compatibility
with existing clients.
Simple change (no body needed):
fix: resolve typo in login error message
Complex change (body required):
fix(auth): resolve race condition in token refresh
The concurrent refresh issue occurred when multiple tabs
tried to refresh expired tokens simultaneously. Now uses
mutex lock to serialize refresh requests.
Fixes #456
When referencing issues or PRs:
fix: resolve race condition in worker
The concurrent access issue occurred when multiple workers tried to
update the same cache entry simultaneously.
Fixes #456
Never add Co-Authored-By footer for AI agents (Claude Code, GitHub Copilot, OpenAI Codex, Cursor, etc.).
AI is a tool, not a co-author. The human takes full ownership of every commit in their repository.
feat: add user dashboard
Implement dashboard with analytics widgets and real-time updates.
All commits are attributed to the human user, regardless of AI assistance. This ensures:
Only use Co-Authored-By for human collaborators:
feat: add user dashboard
Co-Authored-By: Jane Doe <[email protected]>
npx claudepluginhub christophe1997/agent-extentions --plugin agentic-docProvides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.