claude -setup
Easy setup script for Anthropic Claude Model Context Protocol (MCP) servers on Windows
claude mcp add --transport stdio patruff-claude-mcp-setup python setup_mcp.py
How to use
This MCP server runs the Claude setup script to deploy and configure a desktop automation/Claude-based robot setup on Windows. Running the script will install and initialize multiple MCP units (including Gmail/Drive automation, file-system helpers, and other Claude-related components), perform OAuth authentication for Google services, and prepare a local configuration for Claude’s brain on your machine. You can control its behavior with a couple of command line options to streamline authentication and prompts. After deployment, the script stores credentials and configuration files in user directories and sets up a private workspace for Google Drive and Gmail access as part of the robot army.
How to install
Prerequisites:
- Python 3.x installed on Windows
- Git available to clone the repository (optional if you download the script directly)
- Google Cloud project with Gmail/Drive APIs enabled and OAuth credentials prepared (as described in the README)
Step-by-step:
- Clone the MCP repository or download the setup script: git clone <repository-url> cd <repo-directory>
- Ensure Python dependencies are installed (if a requirements file exists at project root): python -m pip install -r requirements.txt
- Place the Google OAuth credentials file in the same directory as the setup script and rename it to gcp-oauth.keys.json (as instructed in the README).
- Run the setup script to deploy Claude:
python setup_mcp.py
- To skip prompts and authentication (stealth mode), use: python setup_mcp.py --skip-prompts --skip-auth
- Verify output and check that the configuration files and credentials were created in the expected user directories.
Notes:
- The script relies on Google Cloud OAuth scopes listed in the README; ensure your OAuth consent screen includes needed scopes and that you’re listed as a test user if required.
Additional notes
Tips and common issues:
- Ensure the gcp-oauth.keys.json file is in the same directory as setup_mcp.py before first run.
- If authentication fails, try re-running with --skip-prompts to bypass API key prompts and verify that APIs are enabled for your Google Cloud project.
- The environment variables for API keys (GIT_PAT_TOKEN, REPLICATE_API_TOKEN, BRAVE_API_KEY, TODOIST_API_TOKEN) may be required by other MCP units; set them in your shell or .env as described in the README.
- The script stores credentials in user directories (e.g., .gmail-server-credentials.json, .gdrive-server-credentials.json). Protect these files as they grant access to Gmail/Drive accounts.
- If you need to reauthenticate, you may need to remove old credential files and re-run the script.
- If you encounter a specific module or path issue, ensure you’re running the command from the repository root and that Python and dependencies are properly installed.
Related MCP Servers
mcp-vegalite
MCP server from isaacwasserman/mcp-vegalite-server
github-chat
A Model Context Protocol (MCP) for analyzing and querying GitHub repositories using the GitHub Chat API.
nautex
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
pagerduty
PagerDuty's official local MCP (Model Context Protocol) server which provides tools to interact with your PagerDuty account directly from your MCP-enabled client.
futu-stock
mcp server for futuniuniu stock
mcp -boilerplate
Boilerplate using one of the 'better' ways to build MCP Servers. Written using FastMCP