Autonomous Python Template

Describe what you want to build. Let agents build it.

A self-building Python ML/Data Science template powered by Claude Code. Clone, describe your project, and watch it come to life.

Quick Start

1. Clone this template

git clone https://github.com/ai-enhanced-engineer/aut-python-template.git my-project
cd my-project

2. Set up environment

just init

3. Start Claude Code

claude

4. Describe your project

> I want to build a RAG system for legal documents

That's it. Describe what you want; agents help you build it.

Example Session

You: I want to build a fraud detection API

Claude: I'll help you build a fraud detection API. Let me understand your requirements:
        - What data sources will you use? (transactions, user behavior, etc.)
        - Any specific ML approach preference? (rule-based, XGBoost, neural network)
        - What's your target latency for predictions?

You: Transaction data from a PostgreSQL database. XGBoost for now, we can upgrade later.
     Need sub-100ms latency.

Claude: Great, let me set up the project structure:
        1. Creating data pipeline for PostgreSQL ingestion
        2. Setting up XGBoost model with feature engineering
        3. Building FastAPI endpoint with async handlers
        4. Adding monitoring and logging

        [Creates files, writes tests, validates...]

You: Add a batch prediction endpoint too

Claude: Adding batch prediction endpoint with:
        - CSV upload support
        - Background processing with status tracking
        - Results stored in PostgreSQL

        [Implements, tests, validates...]

What's Included

This template comes with production-ready infrastructure:

Production Logging

• Structured JSON logging with structlog
• Correlation ID tracking across requests
• Dual-mode: human-readable (dev) / JSON (prod)

Testing Infrastructure

• pytest with markers (unit, functional, integration)
• 80% coverage requirement
• Pre-commit hooks for quality gates

CI/CD Pipeline

• GitHub Actions workflows
• Semantic versioning with auto-release
• Format → Lint → Type-check → Test pipeline

Modern Python Tooling

• Python 3.12+
• Type hints throughout
• Pydantic for data validation
• uv for fast dependency management
• Ruff + Black for formatting/linting
• mypy (strict mode) for type checking

Project Structure

my-project/
├── src/                       # Python source code
│   ├── __init__.py
│   └── logging.py             # Production logging system
├── tests/                     # Test suite
│   └── test_logging.py        # 21+ logging tests as examples
├── .claude/                   # Claude Code settings
├── .github/workflows/         # CI/CD pipelines
│   └── ci.yml
├── CLAUDE.md                  # Development standards
├── pyproject.toml             # Project configuration
├── justfile                   # Automation commands
├── ADR.md                     # Architecture decisions
└── .pre-commit-config.yaml    # Git hooks

Development Commands

just                   # Show all available commands

# Environment
just init              # Complete development setup
just sync              # Update dependencies
just clean-env         # Reset environment

# Code Quality
just format            # Auto-format code
just lint              # Fix linting issues
just type-check        # Validate type hints
just validate-branch   # Run all checks (required before commits)

# Testing
just test              # Standard test suite
just test-unit         # Fast unit tests
just test-functional   # Feature tests
just test-integration  # Integration tests
just test-all          # Complete test suite

The Production-First Philosophy

This template embodies the principle that production AI requires engineering discipline:

• 90% infrastructure, 10% model code: Most production AI is validation, monitoring, error handling, and cost controls—not algorithms
• Reliability over novelty: Production systems must work consistently, not just impressively
• Plan for failure: Every external call needs error handling; every assumption needs validation

Who Should Use This

Teams Starting ML/Data Projects

Stop reinventing infrastructure. Describe your project and get a production-ready foundation.

Senior Engineers New to ML

Get the safety rails you're accustomed to in production systems while learning ML concepts.

Technical Leaders

Give your team a consistent, production-ready starting point that embodies engineering best practices.

Learn More

Production AI Engineering

• A Production-First Approach to AI Engineering
• Google's Rules for ML
• Hidden Technical Debt in ML Systems

Technologies

• FastAPI - Modern Python web framework
• Pydantic - Data validation
• structlog - Structured logging
• uv - Fast Python package management

Contributing

When contributing, prioritize:

1. Reliability over features
2. Simplicity over cleverness
3. Documentation over assumptions
4. Tests over trust

License

Apache License 2.0 - See LICENSE file.

---

"Describe what you want. Let agents build it."