mcp-server-weread
An MCP server for WeChat Reading: your AI agent gets your bookshelf, notes, highlights and top reviews
- 584 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- WEREAD_COOKIE is your WeRead account session: while the cookie is valid, the server reads the account's data with your rights
- With CookieCloud the cookies are synced through a server; the documentation names cc.chenge.ink, which you can replace with your own
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add mcp/mcp-server-wereadDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- WEREAD_COOKIE is your WeRead account session: while the cookie is valid, the server reads the account's data with your rights
- With CookieCloud the cookies are synced through a server; the documentation names cc.chenge.ink, which you can replace with your own
This is third-party code. Review the repository files before installing.
What it does
mcp-server-weread connects the Chinese e-book service WeChat Reading (微信读书, WeRead) to MCP clients. The client receives your account's data: the bookshelf, highlights and notes on books, and top reader reviews. The server has four tools. `get_bookshelf` returns every book on the shelf with title, author, translator and category. `search_books` searches by keyword among the books on your shelf, with fuzzy or exact matching, optional details and a cap on the number of results. `get_book_notes_and_highlights` returns the highlights and notes for a book, grouped by chapter and filtered by highlight style if needed. `get_book_best_reviews` shows top reviews with rating, like count and reviewer data, with paging. The documentation gives a sample dialogue: you ask to show psychology books, the agent finds them on the shelf and then builds a chapter-by-chapter digest of the notes for one of them.
Who it is for. For readers who keep their library and notes in WeRead and want to work with them with an AI agent over MCP.
Good fit when
- You want a book digest built from your highlights and notes, laid out by chapter
- You want to find books on a topic, such as psychology, on your shelf
- You want to look at top reviews of a book with ratings and like counts, paging through them
- You need an overview of what you have read: a list of books with authors, translators and categories
- You need only the highlights of a certain style out of your notes
Not a fit when
- You need a search across the whole WeRead catalog: `search_books` looks only among the books on your shelf
- You need access to the text of the books themselves: the server returns the shelf, notes, highlights and reviews
- You need to write notes back to the service: all four tools only read data
- You are not ready to give the server your WeRead account cookie: it signs in only with a cookie, pasted by hand or synced through CookieCloud
Example request
Show my WeRead shelf and find books on psychology, then build a chapter-by-chapter digest of highlights and notes for one of themMCP
- Transport
- stdio
- Authentication
- key is optional
Environment variables
WEREAD_COOKIEsecret- Your session cookie on weread.qq.com. Needed when CookieCloud is not configured.
CC_URL- The address of the CookieCloud server, your own or a shared one.
CC_IDsecret- The CookieCloud user identifier (UUID).
CC_PASSWORDsecret- The CookieCloud password that the extension shows after automatic generation.
Limitations
It needs Node.js 16 or newer, a WeRead account and a valid cookie. Cookies expire often, which is why the author added CookieCloud sync: once configured, the server takes the cookie from there and uses `WEREAD_COOKIE` as a fallback.
The documentation is written in Chinese only, with no English or Russian instructions, and its links lead to Chinese guide articles. Availability of WeRead from Russia is not covered in the repository. The repository was last updated in May 2025 and the package version in package.json stays at 0.2.2.
How to disable. Remove the mcp-server-weread block from the client configuration and restart it. If you installed the package globally, run `npm uninstall -g mcp-server-weread`. In the browser you can disable the CookieCloud extension.
FAQ
How do I connect WeRead to Claude Desktop?
Open the MCP settings in Claude Desktop and add the mcp-server-weread server with the command npx and the arguments -y mcp-server-weread. In the env block put the CookieCloud details (CC_URL, CC_ID, CC_PASSWORD) or the single WEREAD_COOKIE variable. You can also install the package globally with `npm install -g mcp-server-weread`, in which case the configuration uses the command mcp-server-weread.
How do I get the cookie for WeRead?
Sign in to the web version at weread.qq.com in Chrome, open developer tools with F12, go to the Network tab and reload the page. Find a request to weread.qq.com, copy the full value of the Cookie header and paste it into the WEREAD_COOKIE variable in the MCP configuration or into the .env file.
What is CookieCloud and why does the server use it?
WeRead cookies expire often and have to be refreshed by hand. CookieCloud is an open extension for Edge and Chrome that syncs cookies between the browser and a server. In the extension you set the server address (cc.chenge.ink by default, or your own), enter weread as the domain keyword and run a manual sync. Then you set CC_URL, CC_ID and CC_PASSWORD in the configuration.
Which tools does mcp-server-weread have?
Four: get_bookshelf for the whole shelf, search_books for searching the shelf, get_book_notes_and_highlights for highlights and notes by chapter and get_book_best_reviews for top reviews. All of them only read data.
Which clients does the server work with?
The documentation shows setup for Claude Desktop, and the author's guide describes connecting through Cursor. Any MCP-capable client that can run the npx command and accept environment variables will do.
Discussion
No questions or reviews yet.
Related
MCP serversAn MCP server for your personal WhatsApp account: search and read chats, contacts, send messages and files through an agent
MCP Telegram for reading: the agent gets your account's dialog list and unread messages over MTProto and has no way to write on your behalf
An MCP server for macOS: an agent reads, searches and sends messages in the Messages app on your Mac
WEEEK MCP with a read/write split
weeek-mcp-server
A 12-tool WEEEK MCP server split into 7 read and 5 write tools, so an MCP client can auto-approve the safe ones
WEEEK MCP with confirm-gated writes
weeek-mcp
A WEEEK MCP server that is read-only by default: tasks, attachments and comments via a browser session, writes are opt-in
An MCP server for searching Chinese train tickets on 12306: availability, fares, stations, transfers and stops
README in short
The Chinese document starts with the purpose and a feature list, then lists the four tools with their parameters. Requirements, links to two guides and two ways to integrate with Claude Desktop follow: through npx and through a global install. It describes CookieCloud setup, getting the cookie by hand through developer tools and four sample dialogues with the model. The end collects links to the author's neighboring projects.