{
  "markdown": "# EndNote MCP\n\n[![Tests](https://github.com/gokmengokhan/endnote-mcp/actions/workflows/test.yml/badge.svg?branch=main)](https://github.com/gokmengokhan/endnote-mcp/actions/workflows/test.yml)\n[![PyPI](https://img.shields.io/pypi/v/endnote-mcp)](https://pypi.org/project/endnote-mcp/)\n[![Python](https://img.shields.io/pypi/pyversions/endnote-mcp)](https://pypi.org/project/endnote-mcp/)\n[![License](https://img.shields.io/pypi/l/endnote-mcp)](https://github.com/gokmengokhan/endnote-mcp/blob/main/LICENSE)\n\n<!-- mcp-name: io.github.gokmengokhan/endnote-mcp -->\n\nConnect your EndNote reference library to Claude AI. Search references, read PDFs, format citations, find related papers, and generate bibliographies — all directly in Claude Desktop conversations.\n\n## What It Does\n\nOnce set up, you can ask Claude things like:\n\n- *\"Search my library for social capital and Bourdieu\"*\n- *\"Find papers about how organisations deal with uncertainty\"* (semantic search)\n- *\"Find references related to paper #3844\"*\n- *\"Give me the APA citation for reference #1234\"*\n- *\"Generate a bibliography for references 12, 45, 78, 102\"*\n- *\"Export references 12, 45, 78 as BibTeX\"*\n- *\"Read pages 5-7 from that Smith et al. paper\"*\n\nClaude searches your **local** library — nothing is uploaded to the cloud beyond the normal conversation.\n\n## How It Works\n\n```\nEndNote Library → XML Export → endnote-mcp index → SQLite Database (FTS5 + Embeddings)\n                                                          ↕\n                                   Claude Desktop ← MCP Server (12 tools)\n```\n\nYour references and PDF text are indexed into a local SQLite database with full-text search. Optionally, AI embeddings enable semantic search — finding papers by meaning, not just keywords. When semantic dependencies are installed, new references are automatically embedded during indexing. Claude connects through the [Model Context Protocol (MCP)](https://modelcontextprotocol.io/).\n\n## Requirements\n\n- **EndNote 20+** (any edition)\n- **Claude Desktop** app\n- **Python 3.10+**\n- **uv** (recommended) or pip\n\n## Quick Start\n\n### 1. Install\n\n```bash\n# With uv (recommended)\nuv tool install endnote-mcp\n\n# Or with pip\npip install endnote-mcp\n```\n\n### 2. Export your library from EndNote\n\nIn EndNote: **File → Export** → choose **XML** format → save to a convenient location (e.g., Desktop).\n\n### 3. Run the setup wizard\n\n```bash\nendnote-mcp setup\n```\n\nThe wizard will:\n- Auto-detect your XML export and PDF directory\n- Create the configuration\n- Index your library\n- Configure Claude Desktop automatically\n\n### 4. Restart Claude Desktop\n\nQuit and reopen Claude Desktop. You'll see \"EndNote Library\" in your MCP connectors.\n\nThat's it. Start asking Claude about your references.\n\n## Semantic Search (Optional)\n\nEnable meaning-based search that finds references even when they use different terminology than your query. For example, searching *\"how companies prepare for uncertain futures\"* finds papers on scenario planning and strategic foresight.\n\n```bash\n# Install semantic search dependencies\npip install endnote-mcp[semantic]\n\n# Generate embeddings (~3 min for 4,000 references)\nendnote-mcp embed\n```\n\nThis uses the lightweight [all-MiniLM-L6-v2](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) model locally — no API keys needed, everything stays on your machine.\n\n## Commands\n\n| Command | What It Does |\n|---------|-------------|\n| `endnote-mcp setup` | Interactive setup wizard |\n| `endnote-mcp index` | Re-index after adding new references (incremental) |\n| `endnote-mcp index --full` | Full re-index from scratch |\n| `endnote-mcp index --skip-pdfs` | Index metadata only (fast, ~1 sec) |\n| `endnote-mcp index --embed` | Re-index and generate embeddings |\n| `endnote-mcp embed` | Generate semantic search embeddings |\n| `endnote-mcp embed --full` | Regenerate all embeddings from scratch |\n| `endnote-mcp status` | Show index statistics |\n| `endnote-mcp install` | Add to Claude Desktop config |\n| `endnote-mcp serve` | Start MCP server (used by Claude Desktop automatically) |\n\n## Tools Available to Claude\n\n| Tool | Description |\n|------|-------------|\n| `search_references` | Search by author, title, year, keywords, abstract (BM25 ranked, with DOI links) |\n| `search_fulltext` | Search inside PDF content — find concepts, quotes, methods |\n| `search_library` | Combined metadata + PDF search in one call |\n| `search_semantic` | Search by meaning using AI embeddings (requires `endnote-mcp[semantic]`) |\n| `get_reference_details` | Full metadata for a reference (abstract, keywords, DOI, etc.) |\n| `get_citation` | Format as APA 7th, Harvard, Vancouver, Chicago, or IEEE |\n| `get_bibtex` | Export references as BibTeX entries for LaTeX |\n| `get_bibliography` | Generate a formatted bibliography for multiple references |\n| `find_related` | Find similar references (uses embeddings when available) |\n| `read_pdf_section` | Read specific pages from a PDF attachment |\n| `list_references_by_topic` | Broad topic-based listing |\n| `rebuild_index` | Re-index after updating your EndNote library |\n\n## Adding New References\n\nWhen you add new references to your EndNote library:\n\n1. **Re-export XML** from EndNote (overwrite the same file)\n2. Either:\n   - Run `endnote-mcp index` from a terminal, **or**\n   - Ask Claude: *\"Rebuild my library index\"*\n\nIndexing is **incremental** — it only processes new references and PDFs, not the entire library again. If semantic search is installed, new references are automatically embedded.\n\n## Performance\n\n| Operation | Time (4,000 references) |\n|-----------|------------------------|\n| Metadata indexing | ~1 second |\n| PDF extraction (first time) | ~1 min per 100 PDFs |\n| PDF extraction (incremental) | Only new PDFs |\n| Embedding generation | ~3 minutes |\n| Keyword search | < 50 ms |\n| Semantic search | < 200 ms |\n\n## Configuration\n\nConfig is stored at:\n- **macOS**: `~/Library/Application Support/endnote-mcp/config.yaml`\n- **Windows**: `%APPDATA%/endnote-mcp/config.yaml`\n- **Linux**: `~/.config/endnote-mcp/config.yaml`\n\n```yaml\nendnote_xml: /path/to/your/library.xml\npdf_dir: /path/to/your/Library.Data/PDF\ndb_path: /path/to/library.db    # auto-set by setup\nmax_pdf_pages: 30                # max pages per read request\nsearch_notes: false              # include the EndNote Notes field in searches\n```\n\n### Notes fields\n\nEndNote's **Research Notes** — where you keep your own summaries and judgments —\nis indexed and searched by default, ranked just below keywords and above the\nabstract.\n\nThe plain **Notes** field is indexed but excluded from matching, because\nreference managers use it as a dumping ground on import: author affiliations,\nemail addresses, thesaurus terms. Set `search_notes: true` to include it. The\nsetting takes effect immediately — no re-index needed.\n\n## Citation Styles\n\nFive built-in styles:\n\n- **APA 7th** — `get_citation(rec_number=42, style=\"apa7\")`\n- **Harvard** — `style=\"harvard\"`\n- **Vancouver** — `style=\"vancouver\"`\n- **Chicago** (Author-Date, 17th ed.) — `style=\"chicago\"`\n- **IEEE** — `style=\"ieee\"`\n\nUse `get_bibliography` to format multiple references at once, sorted by author or year.\n\n## BibTeX Export\n\nExport references as BibTeX entries for use in LaTeX:\n\n- Ask Claude: *\"Export references 42 and 100 as BibTeX\"*\n- Or use the `get_bibtex` tool directly with comma-separated record numbers\n\nGenerates complete entries with proper entry types (`@article`, `@book`, `@inproceedings`, etc.), cite keys, and all available fields.\n\n## Troubleshooting\n\n**\"No configuration found\"** — Run `endnote-mcp setup`\n\n**\"XML file not found\"** — Re-export from EndNote: File → Export → XML format\n\n**\"PDF not found\"** — Check that `pdf_dir` in your config points to the correct `.Data/PDF` directory\n\n**Search returns no results** — Run `endnote-mcp index` to rebuild the database\n\n**Claude Desktop doesn't show the tool** — Run `endnote-mcp install`, then restart Claude Desktop\n\n**\"Semantic search is not available\"** — Run `pip install endnote-mcp[semantic]` then `endnote-mcp embed`\n\n## Citing This Software\n\nIf you use this tool in your research, please cite it:\n\n> Gokmen, G. (2026). *EndNote MCP: Connecting EndNote Reference Libraries to Claude AI* (Version 1.4.5) [Computer software]. https://doi.org/10.5281/zenodo.18617546\n\nOr use the \"Cite this repository\" button on GitHub for BibTeX/APA formats.\n\n## License\n\nAGPL-3.0 — free to use, modify, and distribute. See [LICENSE](LICENSE) for details.\n",
  "bytes": 8473,
  "sha": "574286d9af8db66c148b6ace1820b74f4757a80288409ecbbf11f52683920031",
  "repo_slug": "gokmengokhan/endnote-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_gokmengokhan_endnote_mcp_97b68446/readme"
}