Lyceum

Lyceum

An MCP server that enables querying and managing Calibre libraries via chat by interacting with the Calibre content server over HTTP. It allows users to search for books, update metadata, manage authors and tags, and handle book file uploads or conversions.

Category
访问服务器

README

Lyceum

An MCP server for querying and managing a Calibre ebook library via chat. Works with claude.ai and Claude Code. Talks to Calibre's built-in content server over HTTP — no direct database access or CLI tools needed.

Setup

npm install

Requires a running Calibre content server. You can start one from the Calibre GUI (Preferences > Sharing over the net) or from the command line:

calibre-server /path/to/library

Environment Variables

Variable Required Default Description
AUTH_PASSWORD Yes Password for the OAuth authorization page
CALIBRE_SERVER_URL No http://localhost:8080 URL of the Calibre content server
CALIBRE_LIBRARY_ID No Library ID (only needed for multi-library setups)
CALIBRE_USERNAME No Username for Calibre content server (Digest auth)
CALIBRE_PASSWORD No Password for Calibre content server (Digest auth)
BASE_URL No http://localhost:3000 Public URL of this server (used for OAuth redirects and signed URLs)
AUTH_STATE_FILE No /data/auth-state.json Path to persist OAuth state (clients, tokens) across restarts
PORT No 3000 Port to listen on

Running

AUTH_PASSWORD=your-secret npm run dev

For production:

AUTH_PASSWORD=your-secret \
  CALIBRE_SERVER_URL=http://calibre:8080 \
  BASE_URL=https://lyceum.yourdomain.com \
  npm start

Container

Pre-built images are available from GitHub Container Registry:

podman pull ghcr.io/matthewp/lyceum:latest

The container exposes port 3000. Mount a volume at /data to persist OAuth state across restarts:

podman run -d \
  -p 3009:3000 \
  -v lyceum-data:/data \
  -e AUTH_PASSWORD=your-secret \
  -e CALIBRE_SERVER_URL=http://calibre:8080 \
  -e BASE_URL=https://lyceum.yourdomain.com \
  ghcr.io/matthewp/lyceum:latest

To build from source instead, the included Containerfile uses node:24-slim. Node 24 supports native TypeScript type stripping, so no build step is needed — the source runs directly with --experimental-strip-types.

podman build -t lyceum .

For sensitive values, use podman secrets:

printf 'your-secret' | podman secret create lyceum_auth_password -

Then reference them in a quadlet .container file:

Secret=lyceum_auth_password,type=env,target=AUTH_PASSWORD

MCP Tools

Tool Description
list_books List books sorted by most recently added
get_book Get full details for a book (authors, tags, series, formats, custom columns, etc.)
search_books Search using Calibre's query syntax (e.g. author:Asimov, tag:sci-fi)
list_authors List all authors with book counts
list_tags List all tags with book counts
list_series List all series with book counts
get_download_link Get a signed download URL for a book file (expires in 5 minutes)
get_upload_link Get a signed URL to upload a book via browser (expires in 10 minutes)
set_metadata Update metadata fields on a book (title, authors, tags, custom columns, etc.)
set_cover Set a book's cover image from a URL
fetch_metadata Search Google Books for metadata by title, author, or ISBN
convert_book Convert a book to a different format (e.g. EPUB to PDF)

Connecting to Claude

claude.ai

Go to Settings > Connectors > Add custom connector and enter your server's /mcp URL (e.g. https://lyceum.yourdomain.com/mcp). You'll be prompted to authenticate via the OAuth flow.

Claude Code

claude mcp add --transport http lyceum https://lyceum.yourdomain.com/mcp

Authentication

Lyceum uses OAuth 2.1 with dynamic client registration. When a client connects, it registers automatically, then the user authenticates with the AUTH_PASSWORD. Sessions are persisted to disk so they survive server restarts.

Download and upload links use HMAC-SHA256 signed URLs so they can be opened in a browser without additional authentication.

Communication with the Calibre content server uses HTTP Digest authentication when CALIBRE_USERNAME and CALIBRE_PASSWORD are set.

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选