Overview

Project Gutenberg's library of 78,000+ public-domain books, cataloged through the Gutendex API and served from a Gutenberg content mirror. Search by title, author, topic, or language, fetch full book metadata, and retrieve plain-text content in offset/limit chunks for long works from any MCP client. Runs as a stdio process, a local Streamable HTTP server, or the public hosted endpoint above.

Tools

Capability reference

gutenberg_search_books _tool

• Free-text query matches titles and author names; topic matches subjects/bookshelves separately from query
• Filters: languages (ISO 639-1 codes), author_year_start/author_year_end lifespan range, ids for a batch lookup by known ID list
• sort: popular (default, by download count) or ascending/descending by Gutenberg ID
• Paginated — up to 32 books per page; totalCount and hasMore drive further pages
• Each result carries has_plain_text to pre-filter before calling gutenberg_get_text
• Typed failures: no_results, page_out_of_range, catalog_unavailable

gutenberg_get_book _tool

• Returns the full formats map (MIME type → download URL) — plain text, HTML, EPUB, cover image
• Authors, translators, and editors each carry birth/death years
• has_plain_text confirms a UTF-8 plain-text format exists; media_type distinguishes text books from audio ("Sound")
• summary is the first entry of Gutendex's summaries array — both are null/empty when Gutendex has none
• Typed failures: not_found, catalog_unavailable

gutenberg_get_text _tool

• offset/limit chunking (limit 1–50,000, default 20,000) for works that routinely run 500KB–2MB
• Strips the standard Gutenberg license header and footer before chunking
• Prefers UTF-8 plain text; falls back to an HTML-to-text conversion (sourceFormat reports which)
• Response carries totalChars, length, remainingChars, hasMore — use length, not limit, to compute the next offset, since paragraph-boundary trimming can return slightly less than requested
• provenance carries the Gutenberg ID, title, and license URL for attribution
• Typed failures: not_found, audio_book (refuses media_type "Sound"), no_text_format, offset_out_of_range, text_fetch_failed, catalog_unavailable

gutenberg_browse_popular _tool

• Returns up to 32 titles (limit, default 20) ordered by download count, most popular first
• Optional languages and topic filters, applied together
• totalInCatalog gives full context (e.g. "top 20 of 60,000")
• Enrichment reports whether results were truncated and the download-count ceiling of the least-popular book shown, pointing to gutenberg_search_books to page through the rest
• Typed failures: no_results, catalog_unavailable

Features

Built on @cyanheads/mcp-ts-core: stdio and Streamable HTTP transports, pluggable auth (none / jwt / oauth), swappable storage (in-memory, filesystem, Supabase, Cloudflare KV/R2/D1), structured logging with optional OpenTelemetry tracing.

Gutenberg-specific:

• Catalog search and metadata via Gutendex, an unofficial but stable JSON API over the Gutenberg dataset
• Single-book lookups (gutenberg_get_book, gutenberg_get_text) read from a local SQLite+FTS5 catalog mirror when it holds the record, falling back to the live Gutendex API otherwise
• Full plain-text retrieval from a Project Gutenberg content mirror that permits automated access, with transparent UTF-8/HTML fallback
• Tenant-scoped text caching — book text is fetched once per tenant and served from cache for subsequent chunk reads
• No API key required

Agent-friendly output:

• has_plain_text flag on every search/browse result lets agents pre-filter before attempting text retrieval
• Precise chunking contract on gutenberg_get_text — offset, length, totalChars, remainingChars, hasMore for reliable sequential reads
• provenance and discriminated sourceFormat fields on every text response, for attribution and fidelity awareness
• gutenberg_browse_popular enrichment reports truncation state and the download-count ceiling of omitted results

Getting started

Public Hosted Instance

A public instance is available at https://gutenberg.caseyjhand.com/mcp — no installation required. Point any MCP client at it via Streamable HTTP:

{
  "mcpServers": {
    "gutenberg-mcp-server": {
      "type": "streamable-http",
      "url": "https://gutenberg.caseyjhand.com/mcp"
    }
  }
}

Self-Hosted / Local

No API key required. Add the following to your MCP client configuration file:

{
  "mcpServers": {
    "gutenberg-mcp-server": {
      "type": "stdio",
      "command": "bunx",
      "args": ["@cyanheads/gutenberg-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with npx (no Bun required):

{
  "mcpServers": {
    "gutenberg-mcp-server": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@cyanheads/gutenberg-mcp-server@latest"],
      "env": {
        "MCP_TRANSPORT_TYPE": "stdio",
        "MCP_LOG_LEVEL": "info"
      }
    }
  }
}

Or with Docker:

{
  "mcpServers": {
    "gutenberg-mcp-server": {
      "type": "stdio",
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-e", "MCP_TRANSPORT_TYPE=stdio",
        "ghcr.io/cyanheads/gutenberg-mcp-server:latest"
      ]
    }
  }
}

For Streamable HTTP, set the transport and start the server:

MCP_TRANSPORT_TYPE=http MCP_HTTP_PORT=3010 bun run start:http
# Server listens at http://localhost:3010/mcp

Prerequisites

• Bun v1.3.0 or higher (or Node.js v24+).
• No API key required — Project Gutenberg data is freely available.

Installation

1. Clone the repository:

git clone https://github.com/cyanheads/gutenberg-mcp-server.git

2. Navigate into the directory:

cd gutenberg-mcp-server

3. Install dependencies:

bun install

4. Configure environment:

cp .env.example .env
# edit .env if you need to override any defaults

Configuration

See .env.example for the full list of optional overrides.

Running the server

Local development

• Build and run:

  sh

  # One-time build
  bun run rebuild
  
  # Run the built server
  bun run start:stdio
  # or
  bun run start:http
  

• Run checks and tests:

  sh

  bun run devcheck   # Lint, format, typecheck, security
  bun run test       # Vitest test suite
  bun run lint:mcp   # Validate MCP definitions against spec
  

Docker

docker build -t gutenberg-mcp-server .
docker run --rm -p 3010:3010 gutenberg-mcp-server

The Dockerfile defaults to HTTP transport, stateless session mode, and logs to /var/log/gutenberg-mcp-server. OpenTelemetry peer dependencies are installed by default — build with --build-arg OTEL_ENABLED=false to omit them.

Project structure

Development guide

See CLAUDE.md / AGENTS.md for development guidelines and architectural rules. The short version:

• Handlers throw, framework catches — no try/catch in tool logic
• Use ctx.log for request-scoped logging, ctx.state for tenant-scoped storage
• Register new tools via the entry arrays in src/index.ts
• Wrap external API calls: validate raw → normalize to domain type → return output schema; never fabricate missing fields

Contributing

Issues are welcome. Run checks and tests before submitting:

bun run devcheck
bun run test

Data attribution

Data from Project Gutenberg is in the public domain. Catalog metadata sourced from Gutendex (MIT license).

License

Apache-2.0 — see LICENSE for details.