ai-output-validation
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.
$ npx -y skills add DevelopersGlobal/ai-agent-skills --skill api-design --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/api-designContext preview
The summary Claude sees to decide when to auto-load this skill.
Design stable, versioned, self-documenting APIs. Easy to use correctly, hard to use incorrectly. Apply Hyrum's Law from day one.
name: api-design description: Design stable, versioned, self-documenting APIs. Easy to use correctly, hard to use incorrectly. Apply Hyrum's Law from day one. category: build applies-to: [claude, gemini, cursor, copilot, any] version: 1.0.0
APIs are contracts. Once published, every behavior — documented or not — becomes something users depend on (Hyrum's Law). This skill enforces the discipline of designing APIs that are stable, self-documenting, and difficult to misuse.
1. Write the usage examples before writing the implementation. 2. Ask: Is this easy to use correctly? Is it hard to use incorrectly? 3. Apply the principle of least surprise — the API should do what it looks like it does. 4. Design for the caller, not the implementer.
**Verify:** You can write 3 example usages without looking at the implementation.
5. Every observable behavior of your API will be depended upon by someone. 6. Document what IS and IS NOT guaranteed:
7. Be conservative in what you expose — you can always add, never remove.
**Verify:** Every public field and behavior is either documented as stable or marked as internal.
8. Version from day one: `/api/v1/`, `Content-Type: application/vnd.myapi.v1+json` 9. Breaking changes require a new version. 10. Maintain old versions for at least 6 months with deprecation notices. 11. Additive changes (new optional fields) are non-breaking.
**Verify:** API version is in the URL or headers. Deprecation policy is documented.
12. Every endpoint: purpose, inputs, outputs, error codes — documented. 13. Error messages tell the caller what went wrong AND how to fix it. 14. Schema validation on all inputs with meaningful error messages. 15. OpenAPI/Swagger spec generated (not hand-written).
**Verify:** A new developer can use the API from documentation alone, without reading source code.
| Excuse | Rebuttal | |--------|----------| | "We'll document it later" | Undocumented APIs become black boxes. Document as you build. | | "We can break it, it's internal" | Internal APIs become external. Design them well from the start. | | "Versioning is premature" | Retrofitting versioning into an unversioned API is painful. Start versioned. |
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…
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.
Systematic root cause analysis for production and development bugs. Hypothesis-driven debugging — never guess-and-check.