FastAPI + AI Agent + pytest rules for AI coding agents
Full stack: Python + FastAPI + LLM / agent SDKs (Python) + pytest
Ready-to-commit rules for Python + FastAPI + LLM / agent SDKs (Python) + pytest in every major format: AGENTS.md (Codex, Cursor, Copilot), CLAUDE.md (Claude Code), Cursor .mdc project rules, legacy .cursorrules, Copilot instructions, Windsurf rules and GEMINI.md. An AI agent backend: FastAPI serving LLM calls and tools.
Last reviewed 2026-10-21 · Customise in the generator · Raw: AGENTS.md, CLAUDE.md
The mistake agents make most on this stack
Agents let model output flow straight into shell commands or SQL, and never cap token spend.
Generated files
AGENTS.md
# AGENTS.md - this project
Instructions for AI coding agents working in this repository.
## Project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## Project structure
- `app/main.py` creates the app; one `APIRouter` per domain in `app/routers/`.
- Pydantic schemas in `app/schemas/`, business logic in `app/services/`.
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
## Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
## LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
## pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
## Reference docs
- [FastAPI docs](https://fastapi.tiangolo.com)
- [OpenAI Python SDK](https://github.com/openai/openai-python)
- [Anthropic Python SDK](https://github.com/anthropics/anthropic-sdk-python)
- [pytest docs](https://docs.pytest.org)
CLAUDE.md
# CLAUDE.md - this project
Project instructions for Claude Code.
## Project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## Project structure
- `app/main.py` creates the app; one `APIRouter` per domain in `app/routers/`.
- Pydantic schemas in `app/schemas/`, business logic in `app/services/`.
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
## Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
## LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
## pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
## Reference docs
- [FastAPI docs](https://fastapi.tiangolo.com)
- [OpenAI Python SDK](https://github.com/openai/openai-python)
- [Anthropic Python SDK](https://github.com/anthropics/anthropic-sdk-python)
- [pytest docs](https://docs.pytest.org)
.cursor/rules/project.mdc
---
description: "Core rules for this project"
alwaysApply: true
---
# this project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
.cursor/rules/fastapi.mdc
---
description: "Python + FastAPI conventions: FastAPI service with Pydantic v2 and uv."
globs: **/*.py
alwaysApply: false
---
# Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
Docs: https://fastapi.tiangolo.com
.cursor/rules/llm-python.mdc
---
description: "LLM / agent SDKs (Python) conventions: LLM apps and agents using the OpenAI or Anthropic Python SDKs."
globs: **/*agent*.py, **/llm/**/*.py, **/prompts/**
alwaysApply: false
---
# LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
Docs: https://github.com/openai/openai-python, https://github.com/anthropics/anthropic-sdk-python
.cursor/rules/pytest.mdc
---
description: "pytest conventions: Python tests with pytest."
globs: tests/**/*.py, **/test_*.py
alwaysApply: false
---
# pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
Docs: https://docs.pytest.org
.cursorrules
# Cursor rules - this project
You are an expert Python + FastAPI, LLM / agent SDKs (Python), pytest engineer working on this codebase. Follow these rules.
## Project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## Project structure
- `app/main.py` creates the app; one `APIRouter` per domain in `app/routers/`.
- Pydantic schemas in `app/schemas/`, business logic in `app/services/`.
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
## Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
## LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
## pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
## Reference docs
- [FastAPI docs](https://fastapi.tiangolo.com)
- [OpenAI Python SDK](https://github.com/openai/openai-python)
- [Anthropic Python SDK](https://github.com/anthropics/anthropic-sdk-python)
- [pytest docs](https://docs.pytest.org)
.github/copilot-instructions.md
# Copilot instructions - this project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
.github/instructions/fastapi.instructions.md
---
applyTo: "**/*.py"
---
# Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
.github/instructions/llm-python.instructions.md
---
applyTo: "**/*agent*.py,**/llm/**/*.py,**/prompts/**"
---
# LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
.github/instructions/pytest.instructions.md
---
applyTo: "tests/**/*.py,**/test_*.py"
---
# pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
.windsurf/rules/project.md
---
trigger: always_on
---
# this project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
.windsurf/rules/fastapi.md
---
trigger: glob
globs: **/*.py
---
# Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
.windsurf/rules/llm-python.md
---
trigger: glob
globs: **/*agent*.py, **/llm/**/*.py, **/prompts/**
---
# LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
.windsurf/rules/pytest.md
---
trigger: glob
globs: tests/**/*.py, **/test_*.py
---
# pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
GEMINI.md
# GEMINI.md - this project
Project context for Gemini CLI.
## Project
A Python + FastAPI + LLM / agent SDKs (Python) + pytest project.
Stack: Python + FastAPI, LLM / agent SDKs (Python), pytest.
## Commands
- Install deps: `uv sync`
- Dev server: `uv run fastapi dev`
- Lint and format: `uv run ruff check . && uv run ruff format .`
- Tests: `uv run pytest`
- Tests: `uv run pytest -q`
## Project structure
- `app/main.py` creates the app; one `APIRouter` per domain in `app/routers/`.
- Pydantic schemas in `app/schemas/`, business logic in `app/services/`.
## How to work in this repo
- Read the relevant files before editing. Follow the patterns already in the codebase over these generic rules.
- Make the smallest change that solves the task. Do not refactor or reformat unrelated code.
- Before saying a task is done, run the type-check, lint and tests listed under Commands and fix what you broke.
- If requirements are ambiguous, ask one clear question instead of guessing.
- Never commit secrets, API keys or `.env` files. Use env vars and keep `.env.example` up to date.
- Do not add a new dependency when the standard library or an existing dependency already covers it.
- Ask before adding, removing or upgrading any dependency.
- Ask before changing database schemas, auth, payments or CI configuration.
- Never delete files, drop data or run destructive commands (`rm -rf`, `git push --force`, `DROP TABLE`) without explicit approval.
- Keep each change under ~200 changed lines where possible; split larger work into steps.
## Python + FastAPI
- Type-hint every function. Use `Annotated[..., Depends(...)]` for dependencies.
- Request and response bodies are Pydantic v2 models; set `response_model` so output is filtered.
- Use `async def` only when the endpoint awaits async I/O. Blocking libraries go in plain `def` endpoints.
- Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with `pydantic-settings` from env vars.
- Raise `HTTPException` with a clear `detail` for expected errors.
- Manage dependencies with `uv add <pkg>`; do not hand-edit lock files.
## LLM / agent SDKs (Python)
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
## pytest
- Tests live in `tests/` named `test_*.py`. Use plain `assert`.
- Share setup with fixtures in `conftest.py`; use `@pytest.mark.parametrize` for input tables.
- Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
## Reference docs
- [FastAPI docs](https://fastapi.tiangolo.com)
- [OpenAI Python SDK](https://github.com/openai/openai-python)
- [Anthropic Python SDK](https://github.com/anthropics/anthropic-sdk-python)
- [pytest docs](https://docs.pytest.org)
Commands your agent should run
| Task | Command |
|---|---|
| Install deps | uv sync |
| Dev server | uv run fastapi dev |
| Lint and format | uv run ruff check . && uv run ruff format . |
| Tests | uv run pytest |
| Tests | uv run pytest -q |
What the rules cover
Python + FastAPI
FastAPI service with Pydantic v2 and uv. Scoped to **/*.py in Cursor, Copilot and Windsurf.
- Type-hint every function. Use
Annotated[..., Depends(...)]for dependencies. - Request and response bodies are Pydantic v2 models; set
response_modelso output is filtered. - Use
async defonly when the endpoint awaits async I/O. Blocking libraries go in plaindefendpoints. - Inject DB sessions, settings and the current user with dependencies, not globals.
- Load settings with
pydantic-settingsfrom env vars. - Raise
HTTPExceptionwith a cleardetailfor expected errors. - Manage dependencies with
uv add <pkg>; do not hand-edit lock files.
Source docs: FastAPI docs
LLM / agent SDKs (Python)
LLM apps and agents using the OpenAI or Anthropic Python SDKs. Scoped to **/*agent*.py, **/llm/**/*.py, **/prompts/** in Cursor, Copilot and Windsurf.
- Read API keys from env vars; never hardcode them or log them.
- Keep model names, temperatures and token limits in one settings module.
- Store prompts as versioned files or constants, not scattered f-strings.
- Use structured outputs / JSON schema (Pydantic models) when parsing model output; validate before acting on it.
- Set timeouts and retries with backoff on every API call; handle rate limits gracefully.
- Log token usage and cost per call so spend is visible.
- Agent tools must validate their arguments and must never run shell commands or SQL built from raw model output.
Source docs: OpenAI Python SDK, Anthropic Python SDK
pytest
Python tests with pytest. Scoped to tests/**/*.py, **/test_*.py in Cursor, Copilot and Windsurf.
- Tests live in
tests/namedtest_*.py. Use plainassert. - Share setup with fixtures in
conftest.py; use@pytest.mark.parametrizefor input tables. - Mock external services (HTTP, LLM APIs, payment providers); tests must not hit the network.
- Every bug fix adds a regression test.
Source docs: pytest docs
How to install
- Copy
AGENTS.mdto your repo root. That covers Codex, Cursor, GitHub Copilot and most other agents. - Using Claude Code? Add a
CLAUDE.mdwith the single line@AGENTS.md, or use the standalone CLAUDE.md tab above. - Want rules that only load for certain files in Cursor? Add the
.cursor/rules/*.mdcfiles. - Commit the files so your whole team (and every agent) shares them.
FAQ
Where do I put the AGENTS.md for Python + FastAPI + LLM / agent SDKs (Python) + pytest?
Save it as AGENTS.md in the root of your repository, next to your main config file. Codex, Cursor and GitHub Copilot pick it up automatically. For Claude Code, add a CLAUDE.md containing @AGENTS.md (or use the standalone CLAUDE.md on this page).
What is the most common mistake AI agents make with Python + FastAPI?
Agents let model output flow straight into shell commands or SQL, and never cap token spend.
Which commands should the agent run to check its work?
For this stack: uv sync, uv run fastapi dev, uv run ruff check . && uv run ruff format ., uv run pytest. They are listed in the generated files so the agent runs them before finishing a task.
Do these Python + FastAPI + LLM / agent SDKs (Python) + pytest rules work with Cursor and Claude Code?
Yes. This page includes the same rules as AGENTS.md, CLAUDE.md, Cursor .mdc project rules, a legacy .cursorrules file, GitHub Copilot instructions, Windsurf rules and GEMINI.md.
Can I customise these rules?
Yes. Open this stack in the generator to add or remove libraries, set your project name and package manager, switch beginner explanations on, and download everything as a zip.
Related stacks
Python + FastAPI + SQLAlchemy 2 + Alembic + pytest + Docker
A production-style Python API: FastAPI, SQLAlchemy 2 and Alembic, tested with pytest.
Python + FastAPI + Supabase + pytest
A FastAPI service on Supabase Postgres and auth.
Python (scripts, agents, libraries) + LLM / agent SDKs (Python) + pytest
A Python AI agent or LLM script with uv, ruff and pytest.
Python + Django + pytest + Docker
A Django app tested with pytest.
Python + Django + Tailwind CSS v4 + pytest
Server-rendered Django with Tailwind styling.
Next.js (App Router) + Tailwind CSS v4 + shadcn/ui + Supabase + Zod validation
The default vibe-coding SaaS stack: Next.js App Router, Supabase for auth and Postgres, Tailwind and shadcn/ui for the interface.