Skip to content

Installation Guide

This guide walks you through installing and setting up the AutoDocs MCP Server for different AI clients and development environments.

Quick Start

The fastest way to get started is installing from PyPI:

# Using uv (recommended)
uv tool install autodoc-mcp

# Using pip
pip install autodoc-mcp

For development work or to access the latest features:

# Clone the repository
git clone https://github.com/bradleyfay/autodoc-mcp.git
cd autodoc-mcp

# Install with all development dependencies
uv sync --all-extras

# Verify installation
uv run pytest

MCP Client Configuration

Configure AutoDocs with your preferred AI client:

Claude Code Sessions

For temporary sessions in Claude Code:

# 1. Install the server globally
uv tool install autodoc-mcp

# 2. Start the server (available globally)
autodoc-mcp

# 3. Server provides 8 MCP tools automatically

Session Persistence

In Claude Code sessions, you'll need to install and start the server each time. The global autodoc-mcp command is available after installation.

Claude Desktop

Configure Claude Desktop by editing your claude_desktop_config.json:

Location: ~/Library/Application Support/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "autodoc-mcp": {
      "command": "autodoc-mcp",
      "env": {
        "CACHE_DIR": "~/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO"
      }
    }
  }
}

Location: %APPDATA%/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "autodoc-mcp": {
      "command": "autodoc-mcp",
      "env": {
        "CACHE_DIR": "%USERPROFILE%/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO"
      }
    }
  }
}

Location: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "autodoc-mcp": {
      "command": "autodoc-mcp",
      "env": {
        "CACHE_DIR": "~/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO"
      }
    }
  }
}

Cursor Desktop

Add AutoDocs to Cursor via Settings → Rules for AI → MCP Servers:

{
  "mcpServers": {
    "autodoc-mcp": {
      "command": "autodoc-mcp",
      "env": {
        "CACHE_DIR": "~/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO",
        "MAX_DEPENDENCY_CONTEXT": "8"
      }
    }
  }
}

Continue.dev

For Continue.dev integration, add to your config.json:

{
  "mcpServers": [
    {
      "name": "autodoc-mcp",
      "command": "autodoc-mcp",
      "env": {
        "CACHE_DIR": "~/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO"
      }
    }
  ]
}

Generic MCP Client Configuration

For other MCP clients that support the standard configuration format:

{
  "mcpServers": {
    "autodoc-mcp": {
      "command": "python",
      "args": ["-m", "autodocs_mcp.main"],
      "cwd": "/path/to/autodoc-mcp",
      "env": {
        "CACHE_DIR": "~/.cache/autodoc-mcp",
        "LOG_LEVEL": "INFO",
        "MAX_CONCURRENT": "10",
        "MAX_DEPENDENCY_CONTEXT": "8"
      }
    }
  }
}

Custom Installation Path

If you installed AutoDocs in a custom location or are running from source, adjust the command and args accordingly.

Installation Verification

Test Manual Startup

Verify the installation by starting the server manually:

# Should display the FastMCP startup screen
autodoc-mcp

Expected output: 

[INFO] Starting AutoDocs MCP Server v0.4.2
[INFO] FastMCP server initialized with 8 tools
[INFO] Ready for MCP connections

Test MCP Tools

Once configured with your AI client, test these commands:

  1. Test dependency scanning: 

    Ask your AI: "What packages are available in this project?"
    This should use the scan_dependencies tool.

  2. Test documentation context: 

    Ask your AI: "Tell me about the FastAPI package with its dependencies"
    This should use the get_package_docs_with_context tool.

  3. Test cache status: 

    Ask your AI: "What's the status of the AutoDocs cache?"
    This should use the get_cache_stats tool.

Development Setup

Prerequisites

  • Python 3.8+ (Python 3.11+ recommended)
  • uv package manager (recommended) or pip
  • Git for cloning the repository

Clone and Install

# Clone the repository
git clone https://github.com/bradleyfay/autodoc-mcp.git
cd autodoc-mcp

# Install development dependencies
uv sync --all-extras

# Install pre-commit hooks
uv run pre-commit install

Run Tests

Verify everything works correctly:

# Run all 277 tests
uv run pytest

# Run with coverage report
uv run pytest --cov=src --cov-report=html

# Run linting
uv run ruff check

# Type checking
uv run mypy src

Development Tools

The project includes helpful development scripts:

# Test dependency scanning
uv run python scripts/dev.py test-scan

# Test package documentation fetching
uv run python scripts/dev.py test-docs fastapi ">=0.100.0"

# View cache statistics
uv run python scripts/dev.py cache-stats

# Clear the cache
uv run python scripts/dev.py clear-cache

# Get help
uv run python scripts/dev.py --help

Common Installation Issues

Permission Errors

If you encounter permission errors during installation:

# Use user installation
pip install --user autodoc-mcp

# Or create a virtual environment
python -m venv autodoc-env
source autodoc-env/bin/activate  # On Windows: autodoc-env\Scripts\activate
pip install autodoc-mcp

Command Not Found

If autodoc-mcp command is not found after installation:

  1. Check installation: 

    pip show autodoc-mcp

  2. Check PATH: 

    # Add to your shell profile (.bashrc, .zshrc, etc.)
export PATH="$HOME/.local/bin:$PATH"

  3. Use full path: 

    python -m autodocs_mcp.main

Network Configuration

For corporate networks or proxies:

# Set proxy environment variables
export HTTP_PROXY=http://proxy.company.com:8080
export HTTPS_PROXY=http://proxy.company.com:8080

# Install with proxy
pip install --proxy http://proxy.company.com:8080 autodoc-mcp

Cache Directory Issues

If you have issues with the default cache directory:

# Use custom cache location
export CACHE_DIR="/tmp/autodoc-cache"

# Or specify in MCP configuration
{
  "env": {
    "CACHE_DIR": "/path/to/custom/cache"
  }
}

Next Steps

After installation, you can:

  1. Configure the server - See Configuration Guide
  2. Learn about available tools - See MCP Tools Reference
  3. Troubleshoot issues - See Troubleshooting Guide
  4. Start using AutoDocs - See Getting Started Guide

Installation Complete

You now have AutoDocs MCP Server installed and configured. The server will automatically provide intelligent documentation context to your AI assistant.