This project provides a bridge between Zotero research libraries and AI assistants like Claude, enabling users to semantically search, discuss, and analyze their academic papers. It targets researchers and power users who want to leverage AI for deeper insights from their curated literature.

How It Works

Zotero MCP utilizes the Model Context Protocol (MCP) to integrate with AI assistants. It offers AI-powered semantic search using vector embeddings, supporting local (all-MiniLM-L6-v2) and cloud-based (OpenAI, Gemini) models. The system can automatically sync Zotero's database, extract PDF annotations, and retrieve item metadata and full text.

Quick Start & Requirements

• Install: uv tool install "git+https://github.com/54yyyu/zotero-mcp.git" or pip install git+https://github.com/54yyyu/zotero-mcp.git.
• Setup: Run zotero-mcp setup for auto-configuration with Claude Desktop.
• Prerequisites: Python 3.10+, Zotero 7+ (for local full-text access), Claude Desktop or compatible AI assistant. Better BibTeX plugin for Zotero is recommended for enhanced annotation extraction.
• Docs: Zotero MCP docs

Highlighted Details

• AI-powered semantic search for conceptual discovery.
• Direct PDF annotation extraction and search.
• Local and Web API access methods for flexible integration.
• Supports multiple embedding models (free local, OpenAI, Gemini).

Maintenance & Community

The project is actively maintained by 54yyyu. Further community interaction details are not specified in the README.

Licensing & Compatibility

MIT License. Permissive for commercial use and integration with closed-source applications.

Limitations & Caveats

Some functionality, like tagging or library modifications, may be limited when using the local Zotero JS API; using the web library setup is recommended for full capabilities. The Better BibTeX plugin is recommended for optimal PDF annotation extraction.