MCP-Scheduler

BDWiii/MCP-Scheduler

3.1

If you are the rightful owner of MCP-Scheduler and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

MCP Scheduler is a lightweight server that integrates Google Calendar API requests into MCP tools, enabling efficient scheduling and calendar management.

MCP Scheduler

MCP Scheduler is a lightweight MCP (Model Context Protocol) server that wraps Google Calendar API requests as MCP tools.
It includes a client and a LangGraph React Agent that can reason over the tools to schedule, view, and manage calendar events.

-> this agent is meant to use any self-hosted MCP servers. it's supposed to be used as a simple sub-agent in a bigger architecture.

react

Features

  • MCP Server exposing Google Calendar API as tools
  • MCP Client for interacting with the server
  • LangGraph React Agent reasoning over scheduling actions
  • Simple to run and extend

Requirements

  • Python 3.9+
  • Google Calendar API credentials (OAuth 2.0)
  • Required Python dependencies (see requirements.txt)

Installation

  1. Clone the repository
git clone https://github.com/BDWiii/MCP-Scheduler.git
cd mcp-scheduler
  1. Install dependencies
pip install -r requirements.txt
  1. Set up Google API credentials Follow Google Calendar API setup guide Save your credentials file as credentials.json in the project root

Usage

Run the project with:

python -m main

-> note 1 provide the config file with the absolute path to the server so it works.

-> note 2 This system runs on a local LLM using Ollama.

✅ Using Ollama

Make sure you have Ollama installed and running locally. Then, in config.yaml, set the model name you'd like to use—such as "llama3" or "mistral"—and you're ready to go.

🔁 Using a different LLM If you're not using Ollama, you can easily modify the LLM invocation logic to suit your environment, in the agents/mcp_agent.py

Head over to agents/mcp_agent.py and replace the LLM client (currently set up with ChatOllama) with your preferred model or API (e.g., OpenAI, Claude, or vLLM).

This will:

  • Start the MCP server
  • Connect the MCP client
  • Launch the LangGraph React Agent to handle scheduling tasks

Example Interaction

Once running, the agent can perform actions like:

  • Create a new event in Google Calendar
  • List upcoming events
  • Delete or update events