bookshelf-mcp
MCP server that exposes the Bookshelf (Readarr fork) v1 REST API as tools, allowing an LLM to read and manage authors, books, book files, editions, series, download queue, history, indexers, import lists, custom formats, tags, commands, and system status, with full read/write access and destructive tools flagged.
README
bookshelf-mcp
Part of the arr-mcps collection. MCP server exposing Bookshelf's v1 REST API (a Readarr fork, OpenAPI 3.0.1) as tools, so an LLM can read and manage a Bookshelf instance: authors, books, book files, editions, series, the download queue, history, indexers, import lists, custom formats, tags, commands, system status, and more. Full surface — reads and writes, with destructive tools flagged.
Built with FastMCP.
Getting an API key
Generate one in Bookshelf Settings > General > Security. Auth is the
X-Api-Key header.
Install
Download a wheel from the latest release
and install it as a uv tool (no repo checkout needed):
uv tool install bookshelf_mcp-*.whl
This puts a bookshelf-mcp command on your PATH. Register it with Claude Code:
claude mcp add bookshelf \
--env BOOKSHELF_URL=http://your-bookshelf-host:8787 \
--env BOOKSHELF_API_KEY=<key> \
-- bookshelf-mcp
From source
uv sync
cp .env.example .env # fill in BOOKSHELF_URL and BOOKSHELF_API_KEY
claude mcp add bookshelf \
--env BOOKSHELF_URL=http://your-bookshelf-host:8787 \
--env BOOKSHELF_API_KEY=<key> \
-- uv run --directory /path/to/bookshelf-mcp bookshelf-mcp
Config
| Env var | Required | Default |
|---|---|---|
BOOKSHELF_URL |
yes | - |
BOOKSHELF_API_KEY |
yes* | none (no X-Api-Key header sent if unset) |
* Every API endpoint requires auth; practically you must set it, but the server still starts without one so errors surface from the API rather than at startup.
Tools
15 resource-scoped tools, each covering multiple Bookshelf v1 endpoints
(221 total) via an operation parameter. Call a tool with operation set
to one of its listed operations and an arguments dict matching that
operation's parameters — the tool's own description (visible to your MCP
client) lists every operation, its signature, and a one-line doc. This keeps
the full REST surface available while costing a fraction of the context
budget of registering all 221 endpoints as separate tools.
| Tool | Operations | Kind |
|---|---|---|
bookshelf_profiles_formats |
40 | reads + writes |
bookshelf_media_library |
33 | reads + writes |
bookshelf_config |
27 | reads + writes |
bookshelf_system_commands |
20 | reads + writes |
bookshelf_notifications_metadata |
18 | reads + writes |
bookshelf_download_clients |
16 | reads + writes |
bookshelf_import_lists |
16 | reads + writes |
bookshelf_indexers |
11 | reads + writes |
bookshelf_storage |
9 | reads + writes |
bookshelf_history_blocklist |
7 | reads + writes |
bookshelf_queue |
7 | reads + writes |
bookshelf_tags |
7 | reads + writes |
bookshelf_wanted |
4 | read-only |
bookshelf_release_search |
4 | reads + writes |
bookshelf_calendar |
2 | read-only |
Example: bookshelf_queue(operation="bookshelf_delete_queue", arguments={"id": 42}).
Endpoint-level naming (bookshelf_<verb>_<resource>) is preserved as the
operation value, so the full endpoint list is still discoverable from each
group tool's description at runtime.
Development
make help # list all commands
| Command | Does |
|---|---|
make sync |
uv sync |
make test |
Offline tests - one per endpoint, mocked HTTP |
make test-integration |
Tests against the live instance (needs BOOKSHELF_URL/BOOKSHELF_API_KEY) |
make build |
Build wheel + sdist into dist/ |
make bump-patch / bump-minor / bump-major |
Bump the version in pyproject.toml + uv.lock |
make clean |
Remove build artifacts |
The release workflow (.github/workflows/release.yml) builds and publishes to
Releases whenever a v*
tag is pushed - so the usual flow is make bump-patch, commit, then tag and
push.
The integration suite is read-only by default (GET endpoints only). Set
BOOKSHELF_WRITE_TESTS=1 to also exercise POST/PUT/DELETE against a scratch tag
(created, updated, then deleted). Never run write tests against a production
library.
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。