Ai Scheduler MCP

🚀 ai-scheduler-mcp

An MCP (Model Context Protocol) server integrating Google Tasks and Calendar API, allowing Google Tasks and Calendar features to be used from MCP clients (such as Roo Code).

🚀 Quick Start

Prerequisites

• Docker is installed.
• A Docker network named mcp-network exists.

  # If it doesn't exist, create it with the following command:
  docker network create mcp-network
  

Setup Preparation

1. Repository Preparation: Clone or download this repository.

2. Authentication Information (credentials.json) Preparation:

   • Create an OAuth 2.0 client ID (of type "Desktop app" or "Web app") in the Google Cloud Console and enable the Google Tasks API and Google Calendar API.
   • Download the generated authentication information and save it as credentials.json in the project root directory (ai-scheduler-mcp/).
   • Note: credentials.json is confidential information. Never commit it to the Git repository. (It's already included in .gitignore).

3. Docker Image Building: Use the scripts/ai-scheduler-mcp.sh script to manage the project. For convenience, it's recommended to set an alias (e.g., ai-scheduler-mcp).

   # For example, add the following to your shell configuration file:
   alias ai-scheduler-mcp="/path/to/your/project/ai-scheduler-mcp/scripts/ai-scheduler-mcp.sh"
   

   Then specify the port using an environment variable (default is 3003).

   export PORT=${PORT:-3003}
   

Usage Instructions

• Navigate to the project root directory and run the script:

  cd /path/to/your/project/ai-scheduler-mcp
  ./scripts/ai-scheduler-mcp.sh
  

Command Operations

# Build the image. The default port is 3003. You can specify the --port parameter to change it:
ai_scheduler_mcp build [--port ${PORT}]

# Start the container. By default, it uses credentials.json and environment variable configurations:
ai_scheduler_mcp up [--build] [--env-file .env]

# Stop the container:
ai_scheduler_mcp down

# Restart the container:
ai_scheduler_mcp restart

Additional Information

• Using Docker Directly: You can directly use docker build and docker run commands. Refer to the contents of scripts/ai-scheduler-mcp.sh or Docker documentation for specific operations.
• Inter - Container Connection: Other containers in the same mcp-network can connect via http://ai-scheduler-mcp-server:${PORT}/sse.
• Communication Protocol: This server uses SSE (Server - Sent Events) for communication.