ai-output-validation
Validates, parses, and sanitizes AI-generated outputs before they reach end users or downstream systems. Structured output enforcement, schema validation, and…
Trunk-based development with atomic commits, clean history, and meaningful commit messages. Every commit should be deployable.
$ npx -y skills add DevelopersGlobal/ai-agent-skills --skill git-workflow --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/git-workflowContext preview
The summary Claude sees to decide when to auto-load this skill.
Trunk-based development with atomic commits, clean history, and meaningful commit messages. Every commit should be deployable.
name: git-workflow description: Trunk-based development with atomic commits, clean history, and meaningful commit messages. Every commit should be deployable. category: ship applies-to: [claude, gemini, cursor, copilot, any] version: 1.0.0
Git is not just a backup system — it's a communication tool. Clean git history enables fast debugging (git bisect), clear attribution, and safe reverts. This skill enforces atomic commits, meaningful messages, and trunk-based development.
1. Each commit should represent ONE logical change — not a day's worth of work. 2. A commit should be: independently deployable, independently revertable. 3. Never commit "WIP" or partial implementations.
**Verify:** You could revert this commit without affecting adjacent functionality.
4. Follow Conventional Commits:
type(scope): short summary (max 72 chars) Body: what changed and WHY (not how — the diff shows how). Closes: #issue-number
5. Types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore` 6. The summary is imperative mood: "Add feature" not "Added feature"
**Verify:** Message passes: `feat|fix|docs|...(<scope>): <summary>` format.
7. Work directly on `main` for small changes (<1 day of work). 8. For larger features: short-lived feature branches (max 2 days), frequent merges to main. 9. Never let a branch live more than 3 days without merging or rebasing. 10. Use feature flags for incomplete features, not long-lived branches.
**Verify:** No branch is more than 2 days old without a merge/rebase plan.
11. Before every commit: tests pass, linter passes, no secrets in diff. 12. Use pre-commit hooks to enforce automatically.
| Excuse | Rebuttal | |--------|----------| | "I'll clean up the commits later" | You won't. Clean as you go. | | "The commit message doesn't matter" | It matters in 6 months when you're bisecting a production bug. | | "Feature branches protect main" | Long-lived branches cause merge nightmares. Trunk-based is safer. |
AI agent skills for production grade applications
Validates, parses, and sanitizes AI-generated outputs before they reach end users or downstream systems. Structured output enforcement, schema validation, and…
Design stable, versioned, self-documenting APIs. Easy to use correctly, hard to use incorrectly. Apply Hyrum's Law from day one.
Automated quality gates from commit to production. Every merge to main is potentially shippable. No manual steps in the deployment path.
Get layered, context-aware explanations of unfamiliar code. Understand what it does, why it was written that way, and how to work with it safely.
Structured code review focusing on correctness, security, and maintainability. Correctness before style. Every reviewer comment must be actionable.
Load minimum necessary context into agent context windows. Prevents token bloat, reduces cost, and improves focus. Only load what the current task needs.