ai-output-validation
Validates, parses, and sanitizes AI-generated outputs before they reach end users or downstream systems. Structured output enforcement, schema validation, and…
Test real system boundaries, not mocks of mocks. Integration tests verify that components work together, not that they work in isolation.
$ npx -y skills add DevelopersGlobal/ai-agent-skills --skill integration-testing --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/integration-testingContext preview
The summary Claude sees to decide when to auto-load this skill.
Test real system boundaries, not mocks of mocks. Integration tests verify that components work together, not that they work in isolation.
name: integration-testing description: Test real system boundaries, not mocks of mocks. Integration tests verify that components work together, not that they work in isolation. category: test applies-to: [claude, gemini, cursor, copilot, any] version: 1.0.0
Unit tests verify components work in isolation. Integration tests verify they work together. The gap between these two is where most production bugs hide. This skill ensures integration tests cover the real system boundaries.
1. Map every boundary in the system: code ↔ database, code ↔ external API, service ↔ service, frontend ↔ backend. 2. Rank boundaries by failure impact. 3. Integration tests should cover the top 5 highest-impact boundaries.
**Verify:** Boundary map created. Top 5 prioritized.
4. Use a real test database (not SQLite in-memory if prod uses PostgreSQL). 5. Use contract tests for external APIs — test against a recorded real response. 6. Use test containers for services that would otherwise require mocking. 7. Mock only: third-party services you can't control, slow services with contract tests already in place.
**Verify:** No internal boundaries are mocked in integration tests.
8. Happy path: the system works end-to-end for the primary use case. 9. Failure cases: database down, API returns 500, timeout, malformed response. 10. State verification: after each action, verify the state in the real database.
**Verify:** At least 1 failure case per boundary is tested.
11. Each test cleans up its own state (transactions rolled back, test data deleted). 12. Tests don't share state with each other. 13. Tests can run in any order.
**Verify:** Test suite passes when run in random order.
| Excuse | Rebuttal | |--------|----------| | "Unit tests are enough" | Unit tests with mocks test that your mocks work, not your system. | | "Integration tests are slow" | Slow tests are better than discovering bugs in production. Optimize the setup, not the coverage. | | "We test in staging" | Staging is not a substitute for automated tests. It's too slow and too manual. |
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.