Use when adding an LLM, assistant, or chatbot endpoint to a FastAPI service, or when building, integrating, or testing pydantic-ai agents — defining the Agent, typed dependencies, tools, system prompts, model registry, agent-as-FastAPI-dependency wiring, or agent test fixtures (TestModel, FunctionModel, provider error mapping).
Use when working with PostgreSQL via SQLAlchemy 2.0 async and Alembic — defining models, writing service queries (CRUD, pagination, filtering), generating or debugging migrations, setting up testcontainers-based database isolation in tests, or diagnosing `MissingGreenlet`, `lazy='raise'` / `InvalidRequestError`, or N+1 query problems.
Use when writing, reviewing, or refactoring Python 3.13+ code — type hints, code density, model-first data design, naming, dependency injection, early returns, fail-fast discipline, and architectural principles (KISS, YAGNI, SRP, DRY). Does not cover framework patterns, tool configuration, or test layout — those are owned by the matching domain skill.
Use when writing, reviewing, or planning tests for a FastAPI service, or stabilizing flaky tests — API-level test philosophy, polyfactory data generation, test organization, helper patterns, assertion patterns, FastAPI dependency overrides, and coverage requirements. For PostgreSQL container/fixture setup see `postgres-database`; for pydantic-ai test mocking see `ai-agents`.
Use when creating, editing, splitting, or reviewing `skills/<name>/SKILL.md` files in this repository, or when adding evaluation scenarios for those skills. Applies only to this repository's `skills/` tree; for foreign skill systems use that system's own authoring guide.
Use when bootstrapping or changing Python development tooling — uv dependency management, adding or removing dependencies, ruff formatting/linting, ty type checking, pytest configuration in `pyproject.toml`, pre-commit hooks, optional Makefile wrappers, or CI quality jobs.
Use when building or modifying a FastAPI service without a repository layer — adding routes, services, Pydantic schemas, exception handlers, pydantic-settings configuration, dependency injection, or project structure. For SQLAlchemy models and migrations see `postgres-database`; for pydantic-ai agents see `ai-agents`.
Use when starting a brand-new Python/FastAPI service from zero — "create an app that…", bootstrapping a fresh repo, or scaffolding a new microservice/API that needs tests, linters, CI, and Docker working from the first commit. Greenfield only; never for adding features to an existing project. For tooling changes in an existing repo see `python-tooling`.