What problem does it solve? Writing tests in a multi-service repo means knowing which runner each service uses, where spec files live, and which mocking traps cause false-green or false-red results. This Skill encodes all of that so generated tests match existing conventions instead of silently passing or breaking suites. ## Core Features & Use Cases - Runner routing per service: Maps each service (demo_api_server, demo_api_ui, oauth-mcp, demo_mcp_gateway, langchain_agent, openai_agent, pydantic_agent) to its correct runner — Jest, Vitest, or pytest — and its spec file location. - Convention enforcement: Applies repo rules like the { error } response shape, layer-specific MCP error formats, and normalizeAxiosError assertions. - Trap detection: Flags known pitfalls such as jest moduleNameMapper shims, configStore mock shadowing, StrictMode double-invocation, and pytest-asyncio strict mode silently skipping async tests. - Use Case: When asked to "add tests for the transaction-flows route", the Skill picks Jest with supertest, extends the existing spec file, mocks middleware/auth correctly, and asserts the { error } shape. ## Quick Start Ask the AI to generate tests for a specific function or route in a named service, for example: write unit tests for the authorize-gate middleware in demo_api_server.
