EndNote MCP
EndNote MCP searches an EndNote reference library, reads PDFs and builds citations and bibliographies inside Claude Desktop
- 61 ★ on the repository
- Last commit
- AGPL-3.0
Minimal access
Why this level
- The server reads the local library and PDFs, and only the fragments that enter the Claude conversation leave the machine
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/endnote-mcpDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The server reads the local library and PDFs, and only the fragments that enter the Claude conversation leave the machine
This is third-party code. Review the repository files before installing.
What it does
EndNote MCP indexes an XML export of an EndNote library and the text of PDF attachments into a local SQLite database with full-text search. Through 12 tools Claude searches works by keywords, finds papers by meaning, picks related records, formats citations and exports a selection to BibTeX. With the semantic extra installed, embeddings are computed locally and search follows the meaning of a query. The repository also has an `endnote-library` skill: it requires citing only what the tools return, taking page numbers from the printed page and checking quotes word for word. The library stays on the computer and only the normal conversation leaves it.
Who it is for. Researchers and students who keep a library in EndNote 20 and write with Claude.
Good fit when
- You need works on social capital and Bourdieu from your own library
- A literature review needs papers close in meaning to record 3844
- A reference list for a draft is built from records 12, 45, 78 and 102
- A quote has to be checked against a PDF page inside the library
Not a fit when
- The library is kept outside EndNote, for example in Zotero
- You need papers that are not in your library
- An XML export is not possible
Example request
Find papers in my library on how organisations deal with uncertainty and give me an APA citation for record 1234MCP
- Transport
- stdio
- Authentication
- not required
Limitations
EndNote 20 or newer and an XML export of the library are required. The CLI install needs Python 3.10+ and uv or pip. Semantic search is enabled separately (`pip install endnote-mcp[semantic]`, then `endnote-mcp embed`). The license is AGPL-3.0, keep it in mind when embedding into your own products. The server reads PDFs by the paths from the export, so the attachments folder must be given during setup.
How to disable. Remove the extension in Claude Desktop settings or delete the server entry from the configuration and run `pip uninstall endnote-mcp`. The local index database can be deleted by hand.
FAQ
How do I install EndNote MCP in Claude Desktop?
Download `endnote-mcp.mcpb` from the releases and open it, pointing it at the XML export and the PDF folder. The terminal route is `uv tool install endnote-mcp` followed by `endnote-mcp setup`.
Does the library go to the cloud?
No, search runs on a local database. Only the normal conversation with Claude leaves the machine.
What does the endnote-library skill do?
It makes the agent cite only returned records, take pages from the printed page and check quotes word for word. In Claude Code it installs with `/plugin install endnote --marketplace gokmengokhan/endnote-mcp`.
Does EndNote have to stay installed?
No, a single XML export of the library and the attachments folder are enough. After indexing the server reads the local SQLite database, and library changes are picked up with a new export.
Does semantic search work offline?
After installing the `endnote-mcp[semantic]` extra and running `endnote-mcp embed`, embeddings are computed locally. Without the extra, full-text search remains.
Discussion
No questions or reviews yet.
Related
MCP serversA Zotero plugin with a built-in MCP server: the agent searches the library, reads PDFs and notes, and works with annotations and metadata
BibTeX verification against Crossref, OpenAlex, PubMed and arXiv: a CLI, Python API and MCP server to flag wrong references
A Zotero MCP server: keyword and semantic search, PDF passages with page numbers, CSL citations and adding papers by DOI
The MCP server of the seerai Zotero plugin gives outside agents 46 tools: library and scholarly search, reviews and a workspace.
A local MCP server for arXiv literature work: section-level paper reads, BibTeX from metadata, topic watches
Scholar Search MCP connects a discovered paper to its references and later citations when preparing a literature review.
README in short
The README explains the path from the XML export to the SQLite database and the server, then gives a bundle install for Claude Desktop and a PyPI install. It continues with the plugin and skill for agents, a tool list and common problems.