MCP Open Library
An MCP server for Open Library: the agent searches books and authors and gets covers, photos and data by ISBN, OCLC and LCCN
- 102 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- Tools only read the public Open Library catalog and carry `readOnlyHint`
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/mcp-open-libraryDetects 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.
- Tools only read the public Open Library catalog and carry `readOnlyHint`
This is third-party code. Review the repository files before installing.
What it does
The server gives the agent access to the Open Library catalog of the Internet Archive. Seven tools only read data: `search_books` searches by title, author, subject, place, publisher and ISBN with sorting and paging, `get_book_by_title` and `get_authors_by_name` answer simple queries, `get_author_info` and `get_book_by_id` return detailed records by OLID key, ISBN, LCCN or OCLC. Two tools return image links: `get_book_cover` and `get_author_photo`. Every search returns at most 50 results per call and the total match count `num_found`, then the agent pages through `offset` up to 1000. Tools are marked `readOnlyHint` and `openWorldHint`, so a client may skip the confirmation prompt.
Who it is for. Librarians, editors and book service developers who need the agent for references on editions and authors.
Good fit when
- A bibliography needs the first publication year and publisher checked by ISBN
- A book card needs a cover by OCLC or LCCN identifier
- A subject selection sorted from old editions to new is easier to get through `search_books`
- Author details and a photo are needed by Open Library key
Not a fit when
- Prices, store availability or reviews are needed: Open Library does not hold them
- The full book text is needed: the server returns catalog data and links only
- The source must be Russian: the catalog is international and Russian editions may be less covered
Example request
Find the first editions of Ursula K. Le Guin fantasy novels from old to new and show the publication year and coversMCP
- Transport
- stdio
- Authentication
- not required
Limitations
The server runs through `npx` and needs Node.js; the author tested it on version 22. All data comes from the public Open Library API, so record completeness and freshness depend on the catalog. Paging is limited: up to 50 results per call and an offset up to 1000, so long selections must be split. No keys or environment variables are needed. Covers and photos are returned as image addresses and the server does not download the files.
How to disable. Remove the `mcp-open-library` entry from `mcpServers` in the client config and restart the client.
FAQ
How do I connect MCP Open Library?
Add an entry with the `npx` command and the arguments `-y mcp-open-library` to `claude_desktop_config.json`, then restart the client. Other clients use the same entry shape.
How many results does search return?
Up to 50 per call, 10 by default, and a `num_found` field with the total. Further pages come through `offset`, up to 1000.
Which identifiers can find a book?
ISBN, LCCN, OCLC and OLID through `get_book_by_id`, and the cover by the same identifiers plus ID through `get_book_cover`.
What license applies?
MIT.
Is an Open Library account needed?
No, the server uses the public API and works without keys or sign-in.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server that searches Google and Scholar, reads pages and PDFs and stores findings in a local project knowledge graph
A unified search MCP server: one interface to Tavily, Brave, Kagi, Exa, Linkup, Firecrawl and GitHub search
A research MCP server: searches the web, SEC, courts and patents, reads whole pages and checks that cited sources really exist.
A local MCP server with no API keys: multi-engine web search, page fetching and document reading tuned for LLMs.
Remote Consensus MCP server: Claude Code, Codex and Cursor search 220M peer-reviewed papers and answer with links to the studies
Linkup MCP server: real-time web search, page fetching and long research tasks with sources for an agent
README in short
An overview of features, a quick start without installing via `npx`, publication in the MCP registry and a manual build. Then running and testing in the inspector and a tool list with input and output examples.