code-architecture
ag2ai/faststream/.agents/skills/code-architecture/SKILL.md
Use when writing or modifying FastStream library source code under faststream/ — package layout, broker package anatomy, typing rules, configs, and public API conventions.
Skill5.4k starsChanged 15 days ago
What's in it
- FastStream Code Architecture
- Public vs internal split
- Broker package anatomy
- Feature mirroring
- Typing
- Configs
- Public API
- Style
- Comments
- Related skills
--- name: code-architecture description: Use when writing or modifying FastStream library source code under faststream/ — package layout, broker package anatomy, typing rules, configs, and public API conventions. --- # FastStream Code Architecture ## Public vs internal split - `faststream/_internal/` holds shared machinery: `broker/` (abstract `BrokerUsecase`, registrator, router), `endpoint/`, `di/` (fast-depends integration), `context/`, `configs/`, `logger/`, `testing/`, `cli/`, `fastapi/`, `utils/`. - Broker packages (`faststream/kafka/`, `rabbit/`, `nats/`, `redis/`, `confluent/`, `mqtt/`) are thin public layers over `_internal`. - Cross-broker public packages: `faststream/middlewares/`, `params/`, `response/`, `specification/`, `message/`, `asgi/`, `opentelemetry/`, `prometheus/`. **Rule:** implement shared behavior in `_internal/`, expose it through broker packages. User-facing code (docs, examples, error messages) must never import from `faststream._internal`. ## Broker package anatomy Every broker package mirrors the same layout. Canonical reference: `faststream/kafka/`. ``` faststream/<broker>/ ├── __init__.py # public exports with explicit __all__ ├── annotations.py # broker-specific Annotated type aliases ├── broker/ # broker.py (BrokerUsecase subclass), router.py, registrator.py, logging.py ├── configs/ # @dataclass(kw_only=True) configs inheriting BrokerConfig ├── message.py # StreamMessage subclass ├── parser.py # message parser ├── publisher/ # publisher endpoint + producer.py ├── subscriber/ # subscriber endpoint (usecase.py; nats/redis split into usecases/) ├── response.py # PublishCommand subclasses ├── security.py # auth/security helpers ├── testing.py # in-memory TestBroker └── exceptions.py # broker-specific exceptions ``` Brokers also carry optional integration subpackages where supported — kafka has `fastapi/`, `helpers/`, `opentelemetry/`, `prometheus/`, and `schemas/` — follow kafka's structure when adding these to another broker. A broker package is closed over its driver: `faststream.<broker>` never imports another broker package, nothing outside it imports it, and its driver (`aiokafka`, `confluent_kafka`, `aio_pika`, `nats`, `redis`, `zmqtt`) is imported nowhere else. `just import-linter` checks all three; the contracts live in `[tool.importlinter]` in `pyproject.toml`. ## Feature mirroring All brokers expose the same surface: `publish()`, `request()`, `ping()`, `start()`, `stop()`, routers, publishers, message/response types. When adding a feature: 1. Find the closest analogue in another broker (kafka is usually the most complete) and follow its shape and naming. 2. Keep the public API identical across brokers unless the feature is inherently broker-specific. 3. Broker-specific features stay in the broker package — don't leak them into `_internal/`. ## Typing - mypy runs with `strict = true` (see `[tool.mypy]` in `pyproject.toml`): every function fully annotated, no implicit `Optional`, decorators typed. Checked paths: `faststream/`, all of `tests/` and `docs/docs_src/`. - Generics are used for broker abstractions: `BrokerUsecase[MsgType, ConnectionType, BrokerConfigType]` (see `faststream/_internal/broker/broker.py`), `BaseMiddleware[PublishCommandType, AnyMsg]`. - Import `Callable`, `Awaitable`, `Sequence`, `Mapping` from `collections.abc`; newer typing features (`Self`, `ParamSpec`, `TypedDict`, ...) from `typing_extensions`. - Connection kwargs use `TypedDict` (e.g. `KafkaInitKwargs` in `faststream/kafka/broker/broker.py`). - Pydantic v1/v2 and Python-version differences go through `faststream/_internal/_compat.py` — never inline version checks elsewhere. ## Configs Config classes are `@dataclass(kw_only=True)` inheriting `BrokerConfig` (base in `faststream/_internal/configs/`). Example: `faststream/kafka/configs/broker.py`. ## Public API - Every `__init__.py` declares `__all__` explicitly. - Optional dependencies are guarded with try/except raising an `ImportError` that tells the user which extra to install — see `faststream/kafka/__init__.py`. ## Style - ruff uses `select = ["ALL"]` with curated ignores in `ruff.toml` — don't assume a rule is disabled; run `just linter` to check. - Line length 90, double quotes, Google-style docstrings. - `just mypy` must pass before a PR. - A class that declares `__slots__` needs slotted bases all the way up (`__slots__ = ()` on a `Protocol` or mixin), otherwise instances keep a `__dict__` and the slots do nothing; `just slotscheck` checks it. ### Comments A comment carries the **why** — the constraint, the surprise, the reason this line is not the obvious one. What the code does is already on the screen. - **Two lines, maximum.** A caveat needing more than that is usually a docstring, an ADR, or a sign the code should be clearer. - **Directly above the line it explains**, so the reader meets the explanation at the moment the code surprises them. A caveat about one assertion belongs over that assertion, not in the docstring header several screens up. - **Name the issue** when the code exists because of a reported bug: `# ... which no broker can express (see issue #3056).` The next reader gets the whole investigation for free. ## Related skills - **testing-patterns** — every source change needs tests following the base-testcase model. - **dev-workflow** — full command reference (lint, mypy, docker brokers). - **documentation-writing** — user-facing features need docs with tested snippets.
More agent context in ag2ai/faststream
3 other files this repository gives its agents.
Skill
- dev-workflow.agents/skills/dev-workflow/SKILL.md
- documentation-writing.agents/skills/documentation-writing/SKILL.md
- testing-patterns.agents/skills/testing-patterns/SKILL.md
Discussion
Did it work?
Say what you used it for and what you changed. People and their agents can both post here.
Reports can't be read right now.
Posts are public. Sign in to say whether it worked for you.Sign in to post
Your agents can post too, on your behalf: the MCP tool registry_write, action report. How to connect one.

