Zenodo MCP

A comprehensive toolkit for interacting with Zenodo records through the Model Context Protocol (MCP), providing two distinct implementations for different use cases.

Repository Structure

This repository contains two main implementations:

1. MCP SDK Core (/mcp_sdk_core): A Python-based MCP server implementation designed for integration with Cursor IDE and other MCP-enabled environments.

2. MCP API (/mcp_api): A FastAPI-based service that provides MCP-compatible tools for integration with LLM frameworks like LangChain and LangGraph.

Implementation Differences

MCP SDK Core

The MCP SDK Core implementation is designed for direct integration with MCP-enabled environments like Cursor IDE. It provides:

• Direct MCP Integration: Follows the Model Context Protocol standard developed by Anthropic

• Cursor IDE Compatibility: Seamlessly integrates with Cursor's MCP extension

• Simple Configuration: Managed through a JSON config file

• Unified API: Standardized access to Zenodo resources

This implementation is ideal for developers who want to access Zenodo directly from their development environment without additional middleware.

Learn more about the MCP SDK Core implementation →

MCP API

The MCP API implementation is a FastAPI-based service that provides MCP-compatible tools for integration with LLM frameworks. It offers:

• LangChain Integration: Seamless integration with LangChain agents and tools

• LangGraph Compatibility: Support for LangGraph workflows and graphs

• OpenAI-Compatible API: Can be used with OpenAI-compatible clients

• LibreChat Support: Compatible with LibreChat and similar platforms

• Custom Tool Creation: Extensible architecture for creating custom tools

This implementation is ideal for developers building LLM applications that need to interact with Zenodo as part of a larger workflow.

Learn more about the MCP API implementation →

Features

Both implementations provide access to Zenodo's rich repository of research outputs:

• Search and Retrieve Records: Find and access Zenodo records

• Get Citations: Retrieve citations in various formats (BibTeX, APA, etc.)

• Detect Data Types: Automatically classify Zenodo records

• Access Metadata: Get detailed information about records

• List and Download Files: Browse and download files from records

Getting Started

Choose the implementation that best fits your needs:

For Cursor IDE Integration

# Clone the repository
git clone https://github.com/yourusername/zenodo-mcp.git
cd zenodo-mcp/mcp_sdk_core

# Install dependencies
pip install -r requirements.txt

# Configure Cursor IDE (create mcp.json)
# See mcp_sdk_core/README.md for details

For LLM Framework Integration

# Clone the repository
git clone https://github.com/yourusername/zenodo-mcp.git
cd zenodo-mcp/mcp_api

# Install dependencies
pip install -r requirements.txt

# Set up environment variables
cp .env.example .env
# Edit .env with your Zenodo API token

# Run the API server
uvicorn server.main:app --host 0.0.0.0 --port 8000

Contributing

We welcome contributions to both implementations! Please see the respective README files for contribution guidelines.

License

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