llms.txt is an open proposal for a short Markdown file that lists a project's key docs and links, so AI tools can find them without crawling every page.

Where should llms.txt be deployed?

At /llms.txt on your documentation site, or in your repository root. Tools read it when they fetch it; nothing loads it automatically.

# Python FastAPI & AI Agent Service Architecture

> Production standards for Python 3.12+, FastAPI, async I/O, Pydantic v2 schemas, LangChain/LlamaIndex integration, and defensive error propagation.

## Quick Reference

- **Framework**: FastAPI

- **Language**: Python 3.12+

- **Styling & UI**: None / API Service

- **Database / Persistence**: PostgreSQL / pgvector / Redis

- **Architectural Philosophy**: strict

## Core Architectural Invariants

- **Clean Layered Architecture**: Decouple presentation, business logic, and data access layers via interfaces.

- **Guard Clauses & Early Returns**: Exit early from functions to eliminate deeply nested if-else pyramids.

- **Explicit Result / Tuple Returns**: Return [error, result] or Result<T, E> types for operations that can fail.

- **Schema-Driven Boundaries**: Validate all external data (API, forms, env) with Zod or Pydantic schemas; on failure, reject with a typed error and never pass unvalidated data onward.

## Operational Guardrails & Prohibited Patterns

- **Inspect Before Acting**: Always read and understand existing code and imports before suggesting changes.

- **Minimal Surgical Diffs**: Touch only lines relevant to prompt. Preserve existing comments and formatting.

- **Concise & Direct Tone**: Skip conversational pleasantries. Deliver focused diffs and a short rationale.

- **Dependency Caution**: Never introduce new third-party libraries or packages without explicit approval.

- **Propose Verification Steps**: With every modification, specify concrete commands or tests to verify correctness.

## Mandatory Project Directives

- Never hardcode OpenAI/Anthropic API keys or secrets in source code.

- Always implement streaming responses for LLM text generation endpoints.

- Type annotate all function arguments and return types.

## Implementation Reference

### Preferred Patterns

```python

# Good: Typed Pydantic v2 request with dependency injection

from fastapi import APIRouter, Depends, HTTPException, status

from pydantic import BaseModel, Field

router = APIRouter()

class PromptRequest(BaseModel):

query: str = Field(..., min_length=3, max_length=1000)

temperature: float = Field(0.7, ge=0.0, le=1.0)

@router.post("/v1/generate", response_model=GenerationResponse)

async def generate(req: PromptRequest, svc: AIService = Depends(get_ai_service)):

result = await svc.process_prompt(req.query, req.temperature)

if not result.is_ok:

raise HTTPException(status_code=status.HTTP_502_BAD_GATEWAY, detail=result.error_msg)

return result.payload

```

### Discouraged Anti-Patterns

```python

# Discouraged: Untyped dict params and missing error handling

@app.post("/generate")

def generate(data: dict):

result = client.chat(data["query"])

return result

```

## Verification Commands & Workflows

1. Define all request and response schemas strictly using Pydantic v2 BaseModels with Field validation.

2. Use asynchronous route handlers (async def) for any network I/O, LLM inference calls, or database operations.

3. Manage database sessions and external clients using FastAPI Depends injection.

4. Structure the repository into: api/ (routes), services/ (business logic), core/ (config), models/ (schemas).

5. Catch domain-specific exceptions at the service layer and translate to HTTPException at the route layer.

## Context Documents & Specs

- [System Architecture](ARCHITECTURE.md): Authoritative system design, data flows, and state machine invariants.

- [AI Multi-Agent Spec](AGENTS.md): Universal specification for autonomous coding agent operations.

Backend · Any LLM tool that fetches it

Python FastAPI & Pydantic v2 llms.txt

An llms.txt for a Python FastAPI & Pydantic v2 project: a short Markdown map of what it is and where the key docs and code live.

Path: llms.txt
Loads: Fetched on request
Format: Markdown index (.txt)
Used by: Any LLM tool that fetches it

All llms.txt presets and how llms.txt works

What this file tells the agent

An llms.txt gives language models and agents a short Markdown map of your Python FastAPI & Pydantic v2 project: what it is and where the key docs and code live, so a tool that fetches it can go straight to the right files.

Use strict Pydantic v2 BaseModels with Field validation for all input and output schemas.
Use asynchronous route handlers (`async def`) for database operations and external API requests.
Inject dependencies using FastAPI `Depends` pattern; avoid global state.
Type annotate 100% of function signatures with Python 3.12+ type hints.

Code to avoid, and what to write instead

The rule steers the agent away from the left and toward the right.

avoid

@app.post("/users")
def make_user(data: dict):
    # Missing types, sync blocking query
    return db.query(data["email"])

write instead

from fastapi import APIRouter, Depends, HTTPException
from pydantic import BaseModel, Field

router = APIRouter()

class UserPayload(BaseModel):
    email: str = Field(..., min_length=5)

@router.post("/users")
async def create_user(payload: UserPayload, db = Depends(get_db)):
    return await db.users.create(payload.model_dump())

Install it in one command

01
Go to your project root
The file goes at llms.txt, relative to the root.
02
Download the file
Run the command on the right. It downloads this preset and writes the file.
03
Make sure it gets used
Nothing loads llms.txt automatically. Serve it at your site root, or point agents to it from CLAUDE.md or AGENTS.md.

$ curl -fsSL "https://www.devscratchpad.tech/api/raw/llms-txt/fastapi" -o "llms.txt"

writes llms.txt

path

llms.txt

To write the file from your own code instead, use Read your code or Convert Legacy Rules in the editor above.

More for Python FastAPI & Pydantic v2

Other formats

Other llms.txt presets

Questions

What is llms.txt?: llms.txt is an open proposal for a short Markdown file that lists a project's key docs and links, so AI tools can find them without crawling every page.
Where should llms.txt be deployed?: At /llms.txt on your documentation site, or in your repository root. Tools read it when they fetch it; nothing loads it automatically.

Target Standard & File Format

Identity & Activation Rules

Technology Stack Context

Engineering Philosophy

Architectural & Code Quality Conventions

Agent Behavioral Guardrails

Step-by-Step Workflow Procedures

Target File Location