Bookbrainz

What Bookbrainz does

Bookbrainz exposes the public BookBrainz database through three MCP tools. BookBrainz is a MetaBrainz Foundation project and a sister service to MusicBrainz, focused on open bibliographic data. The integration can find and inspect six entity categories: works, editions, authors, publishers, series, and edition-groups.

Search accepts free-text input and returns matching records with BookBrainz UUIDs, also called bbids. Those identifiers can then be passed to lookup to retrieve a specific record. Browse provides relationship-based lists, such as editions associated with a work or works associated with an author.

The data source is BookBrainz's public REST and JSON API at https://api.bookbrainz.org/. No authentication is required. BookBrainz has a smaller coverage footprint than services such as Open Library or Google Books, but it is intended as the canonical source for its open, structured work and edition records.

How it works

The server communicates with an MCP client using either a hosted HTTP endpoint or a local stdin/stdout process. The hosted option is available at https://gateway.pipeworx.io/bookbrainz/mcp; the local package is published as @pipeworx/mcp-bookbrainz and can be started with npx.

A hosted connection lists the three BookBrainz tools alongside shared Pipeworx gateway tools. These additional tools can route questions to other catalog sources, discover tools, or provide gateway-wide functions. A local stdio connection exposes only the BookBrainz tool set and does not add those shared gateway tools. Both modes use the same BookBrainz source and require no account for the initial requests.

The tools can also be called through the gateway's HTTP tool endpoint. For example, a client can send a JSON POST request to /v1/tools/bookbrainz_search with a query and entity type instead of establishing an MCP session.

Setup and configuration

For a hosted MCP connection, add the gateway URL to the client configuration:

{
  "mcpServers": {
    "bookbrainz": {
      "url": "https://gateway.pipeworx.io/bookbrainz/mcp"
    }
  }
}

For a local, gateway-free setup, configure the client to launch the package with npx:

{
  "mcpServers": {
    "bookbrainz": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-bookbrainz"]
    }
  }
}

Running npx -y @pipeworx/mcp-bookbrainz directly starts the stdio server. The package needs no environment variables or Pipeworx account. The hosted gateway and local process are independent connection choices: use the gateway for shared Pipeworx routing, or use stdio for a narrowly scoped BookBrainz connection.

Tools and capabilities

• search(query, type, limit?, offset?) searches one supported entity type using free text. Supported types are work, edition, author, publisher, series, and edition-group.
• lookup(type, bbid, includes?) retrieves one entity by its BookBrainz UUID. The optional includes argument can request aliases, relationships, and identifiers.
• browse(type, ...) lists entities filtered through relationships to another entity, including editions by work and works by author.

Search results are useful for discovering bbids; lookup is the precise follow-up when an agent needs one known record and selected related resources. Browse is suited to relationship-oriented questions that do not start with a free-text search.

Limitations and notes

This is an independent, unofficial integration and is not affiliated with, endorsed by, or published by the upstream BookBrainz provider. The available catalog is smaller than some broader book databases, so coverage may vary by title, author, or edition. Hosted connections include shared gateway schemas in addition to the BookBrainz tools, while local stdio connections remain limited to the pack's own tools.