docpilot

AI-powered documentation autopilot for Python projects

docpilot automatically generates professional, comprehensive docstrings for your Python code using AI. Say goodbye to manual documentation and hello to intelligent, context-aware docstrings that follow your preferred style guide.

Features

• AI-Powered Generation: Leverage GPT-4, Claude, or local LLMs (Ollama) to generate intelligent, context-aware docstrings
• Multiple Docstring Styles: Full support for Google, NumPy, and Sphinx docstring formats
• Smart Code Analysis: Understands your codebase structure, type hints, and complexity metrics
• Production-Ready CLI: Beautiful terminal UI with progress tracking and detailed reporting
• Flexible Configuration: Configure via TOML files, environment variables, or CLI arguments
• Zero-Cost Option: Use local LLMs (Ollama) for completely free operation
• Batch Processing: Process entire codebases with intelligent file discovery
• Safe by Default: Dry-run mode and diff preview before making changes

Installation

Basic Installation

pip install docpilot

With Cloud LLM Support

# OpenAI and Anthropic support
pip install "docpilot[llm]"

With Local LLM Support

# Ollama support for local, free LLM inference
pip install "docpilot[local]"

Full Installation

# Everything including development tools
pip install "docpilot[all]"

Quick Start

1. Initialize Configuration

docpilot init

This creates a docpilot.toml configuration file with sensible defaults.

2. Set Up Your LLM Provider

Option A: OpenAI

export OPENAI_API_KEY="your-api-key-here"

Option B: Anthropic (Claude)

export ANTHROPIC_API_KEY="your-api-key-here"

Option C: Local LLM (Free)

# Install Ollama from https://ollama.ai
ollama pull llama2

3. Generate Docstrings

# Generate for a single file
docpilot generate mymodule.py

# Generate for entire project
docpilot generate ./src --style google

# Preview changes without modifying files
docpilot generate ./src --dry-run --diff

Usage Examples

Analyze Code Structure

Examine your code without generating documentation:

docpilot analyze ./src --show-complexity --show-patterns

Output:

Analyzing ./src/myproject/utils.py...
✓ Found 15 elements (12 public, 3 private)

Functions:
├── calculate_total (line 10) - complexity: 3
├── validate_input (line 25) - complexity: 5
└── process_data (line 45) - complexity: 8

Generate with Custom Configuration

docpilot generate ./src \
  --provider anthropic \
  --model claude-3-sonnet-20240229 \
  --style numpy \
  --include-private \
  --overwrite

Test LLM Connection

Verify your API credentials before processing:

docpilot test-connection --provider openai

Real-World Example

Before:

def calculate_compound_interest(principal, rate, time, frequency):
    return principal * (1 + rate / frequency) ** (frequency * time)

After (Google Style):

def calculate_compound_interest(principal, rate, time, frequency):
    """Calculate compound interest for a given principal amount.

    This function computes the future value of an investment using the
    compound interest formula: A = P(1 + r/n)^(nt), where interest is
    compounded at regular intervals.

    Args:
        principal (float): The initial investment amount in dollars.
        rate (float): Annual interest rate as a decimal (e.g., 0.05 for 5%).
        time (float): Investment duration in years.
        frequency (int): Number of times interest is compounded per year
            (e.g., 12 for monthly, 4 for quarterly).

    Returns:
        float: The total amount after interest, including the principal.

    Examples:
        >>> calculate_compound_interest(1000, 0.05, 10, 12)
        1647.01

        >>> calculate_compound_interest(5000, 0.03, 5, 4)
        5806.11
    """
    return principal * (1 + rate / frequency) ** (frequency * time)

Configuration

Configuration File

Create docpilot.toml in your project root:

[docpilot]
# Docstring style: google, numpy, sphinx, or auto
style = "google"

# Overwrite existing docstrings
overwrite = false

# Include private elements (with leading underscore)
include_private = false

# Code analysis options
analyze_code = true
calculate_complexity = true
infer_types = true
detect_patterns = true

# Generation options
include_examples = true
max_line_length = 88

# File patterns
file_pattern = "**/*.py"
exclude_patterns = [
    "**/test_*.py",
    "**/*_test.py",
    "**/tests/**",
    "**/__pycache__/**",
]

# LLM settings
llm_provider = "openai"
llm_model = "gpt-3.5-turbo"
llm_temperature = 0.7
llm_max_tokens = 2000
llm_timeout = 30

# Project context (helps generate better docs)
project_name = "My Awesome Project"
project_description = "A Python library for awesome things"

Environment Variables

All configuration can be set via environment variables with the DOCPILOT_ prefix:

export DOCPILOT_STYLE="numpy"
export DOCPILOT_LLM_PROVIDER="anthropic"
export DOCPILOT_LLM_MODEL="claude-3-haiku-20240307"
export DOCPILOT_OVERWRITE="true"

CLI Arguments

CLI arguments override both file and environment configuration:

docpilot generate ./src \
  --style sphinx \
  --provider local \
  --model llama2 \
  --overwrite

Supported Docstring Styles

Google Style (Default)

def example(arg1, arg2):
    """Short description.

    Longer description if needed.

    Args:
        arg1 (int): Description of arg1.
        arg2 (str): Description of arg2.

    Returns:
        bool: Description of return value.

    Raises:
        ValueError: When validation fails.
    """

NumPy Style

def example(arg1, arg2):
    """Short description.

    Longer description if needed.

    Parameters
    ----------
    arg1 : int
        Description of arg1.
    arg2 : str
        Description of arg2.

    Returns
    -------
    bool
        Description of return value.

    Raises
    ------
    ValueError
        When validation fails.
    """

Sphinx Style

def example(arg1, arg2):
    """Short description.

    Longer description if needed.

    :param arg1: Description of arg1.
    :type arg1: int
    :param arg2: Description of arg2.
    :type arg2: str
    :return: Description of return value.
    :rtype: bool
    :raises ValueError: When validation fails.
    """

LLM Providers

OpenAI

Best for: High-quality, consistent docstrings

# Supported models
docpilot generate ./src --provider openai --model gpt-4
docpilot generate ./src --provider openai --model gpt-3.5-turbo

Anthropic (Claude)

Best for: Detailed explanations and complex code

# Supported models
docpilot generate ./src --provider anthropic --model claude-3-opus-20240229
docpilot generate ./src --provider anthropic --model claude-3-sonnet-20240229
docpilot generate ./src --provider anthropic --model claude-3-haiku-20240307

Local (Ollama)

Best for: Privacy, cost-free operation, offline work

# First, pull a model
ollama pull llama2
ollama pull codellama
ollama pull mistral

# Then use it
docpilot generate ./src --provider local --model llama2

Requirements

• Python: 3.9 or higher
• Operating System: Linux, macOS, Windows
• Optional: OpenAI API key, Anthropic API key, or Ollama installation

Advanced Usage

Using in CI/CD

# .github/workflows/docs.yml
name: Generate Documentation

on: [push, pull_request]

jobs:
  generate-docs:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3

      - name: Set up Python
        uses: actions/setup-python@v4
        with:
          python-version: '3.11'

      - name: Install docpilot
        run: pip install "docpilot[llm]"

      - name: Generate docstrings
        env:
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
        run: |
          docpilot generate ./src --dry-run

Integration with pre-commit

# .pre-commit-config.yaml
repos:
  - repo: local
    hooks:
      - id: docpilot
        name: Generate docstrings
        entry: docpilot generate --dry-run
        language: system
        types: [python]

Programmatic Usage

from docpilot.core.generator import DocstringGenerator
from docpilot.llm.base import create_provider, LLMConfig, LLMProvider
from docpilot.core.models import DocstringStyle

# Configure LLM
config = LLMConfig(
    provider=LLMProvider.OPENAI,
    model="gpt-3.5-turbo",
    api_key="your-api-key"
)

# Create generator
llm = create_provider(config)
generator = DocstringGenerator(llm_provider=llm)

# Generate docstrings
import asyncio

async def generate():
    results = await generator.generate_for_file(
        file_path="mymodule.py",
        style=DocstringStyle.GOOGLE,
        include_private=False,
        overwrite_existing=False
    )

    for doc in results:
        print(f"{doc.element_name}: {doc.docstring}")

asyncio.run(generate())

Performance

docpilot is designed for production use with large codebases:

• Parallel Processing: Processes multiple files concurrently
• Smart Caching: Avoids redundant LLM calls
• Rate Limiting: Respects API rate limits automatically
• Incremental Updates: Only processes changed files

Typical Performance:

• Small project (50 functions): ~2-3 minutes with OpenAI
• Medium project (500 functions): ~20-30 minutes with OpenAI
• Large project (5000 functions): ~3-4 hours with OpenAI
• Local LLM: 2-3x slower but free

Troubleshooting

API Key Issues

# Verify API key is set
echo $OPENAI_API_KEY

# Test connection
docpilot test-connection --provider openai

Rate Limiting

docpilot automatically handles rate limits, but you can adjust concurrency:

[docpilot]
llm_timeout = 60  # Increase timeout
llm_max_tokens = 1000  # Reduce token usage

Quality Issues

If generated docstrings aren't meeting your standards:

1. Try a better model: Switch from gpt-3.5-turbo to gpt-4
2. Provide context: Set project_name and project_description in config
3. Lower temperature: Reduce llm_temperature to 0.3 for more focused output
4. Add custom instructions: Use custom_instructions in config

Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

Development Setup

# Clone repository
git clone https://github.com/yourusername/docpilot.git
cd docpilot

# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

# Run type checking
mypy src/docpilot

# Run linting
ruff check src/
black --check src/

Roadmap

• Support for JavaScript/TypeScript
• Visual Studio Code extension
• Documentation website generation
• Custom LLM prompt templates
• Docstring quality scoring
• Automated documentation updates on code changes
• Integration with Sphinx/MkDocs

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• Built with Click for CLI
• Powered by OpenAI, Anthropic, and Ollama
• Terminal UI by Rich

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: Read the Docs

---

If docpilot saves you time, please consider giving it a star on GitHub!