Source Library

Source Library MCP Server

[npm version]

Search, read, and cite 22,000+ rare pre-modern texts (alchemy, Hermetica, Kabbalah, theology, early science) with AI-generated English translations. 12 tools (full-text + semantic search, reading, exact quoting, image search, duplicate detection, feedback, sharing findings back). CLI + MCP server, no API key needed.

Quick Start

Claude Code

claude mcp add source-library -- npx -y @source-library/mcp-server

Claude Desktop

Add to your Claude Desktop config:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{  "mcpServers": {    "source-library": {      "command": "npx",      "args": ["-y", "@source-library/mcp-server"]    }  }}

CLI

# Install globallynpm install -g @source-library/mcp-server
# Or run directlynpx @source-library/mcp-server search "philosopher's stone"

From Source

git clone https://github.com/Embassy-of-the-Free-Mind/sourcelibrary-v2.gitcd sourcelibrary-v2/mcp-servernpm install && npm run buildnpm start

12 Tools

Search & Discovery

search_library

Full-text search across books and page content. Returns matching books and pages with citation URLs.

search_translations

Search inside translated text across the entire library. Find what historical authors wrote about any topic. Returns passage snippets with page numbers, book info, and citation URLs.

search_concept

Semantic passage search using Gemini embeddings. Finds conceptually related passages even when the modern term won't literally appear — e.g. "distributed cognition" maps to passages about active intellect, art of memory, wax-tablet metaphors. Prefer search_translations for literal phrases; use search_concept when the concept matters more than the wording.

search_within_book

Search inside a specific book's pages. Returns matching pages with snippets and citation URLs.

list_books

Browse the collection with filters. Returns title, author, language, year, and translation progress.

Reading

get_book

Detailed book metadata: summary, index stats, chapters, edition info, DOI.

get_book_text

Read a book. Prefer chapter for chapter-at-a-time reading; falls back to from/to page ranges. Each response includes a truncated flag with a truncation_note giving the next from/to values when content was cut by the page-budget limit — agents must check truncated rather than infer end-of-book from pages_returned.

get_quote

Get the exact translated text of a single page for quoting. Returns the verbatim translation, original OCR, and a formatted citation. The response headline is citation_link — the stable sourcelibrary.org/q/… shortlink to present alongside the quote and its page number. Always use this before putting text in quotation marks — never paraphrase from memory. Suggested render: > [quote] then — [Author], [citation.locator]. [citation_link] — the locator is "p. 217 [scan 219]" where the printed page is known, else "p. 219".

Gallery

search_images

Search the visual collection: 50,000+ illustrations extracted from book pages PLUS 23,000+ standalone artworks (paintings, frescoes, prints, sculptures from Met, Rijksmuseum, Wikimedia, NGA). Each result has a source field: gallery (book illustration) or artwork (standalone work).

Curation

check_duplicate

Check whether a book already exists in Source Library before importing. Uses 4-tier matching (source fingerprint, title+author normalization, keyword search, semantic similarity) and returns a confidence level + suggestion. Use this before every import to avoid duplicates.

Feedback

submit_feedback

Submit feedback, bug reports, or feature requests directly to the Source Library team.

share_findings

Share a research dossier back to the Source Library team — a title, optional summary, and an ordered list of citations ({ book_id, page, note }, references not copied text). Like submit_feedback, it goes to the team for review.

CLI

Most tools are also available as a standalone CLI with colored terminal output. (search_concept and check_duplicate are MCP-only for now.)

# Search the collectionsource-library search "Paracelsus" --language=German
# Search inside translationssource-library translations "harmony of the spheres"
# Read a booksource-library text 694f49d3... --from=1 --to=50
# Book detailssource-library book 694f49d3...
# Browse the gallerysource-library images --subject=alchemy --type=emblem
# Submit feedbacksource-library feedback "Great translation of Fludd!" --name="Jane"
# JSON output for pipingsource-library search "alchemy" --json | jq .results

Example Research Prompts

"Search for references to 'prima materia' across the collection. Which authors discuss it, and how do their treatments differ?"

"Read the full translation of Fludd's History of Both Worlds, pages 1-50. Summarize the cosmological framework."

"Find all alchemical emblems depicting the ouroboros. What texts are they from?"

"What does Copernicus say about the Sun's centrality in De Revolutionibus? Find the key passages with citation URLs."

Citation URLs

Every page returned by the tools includes a citation URL:

https://sourcelibrary.org/book/fludd-utriusque?page=57

Published editions include DOIs minted via Zenodo.

Development

npm run dev    # Run with hot reload (tsx)npm run build  # Compile TypeScriptnpm start      # Run compiled versionnpm run cli    # Run CLI in dev mode

License

MIT

Links

• Website: sourcelibrary.org
• Developers: sourcelibrary.org/developers
• GitHub: Embassy-of-the-Free-Mind/sourcelibrary-v2