ChatSum MCP Server
Author: chatmcp
Description: An MCP (Model Context Protocol) server for querying and summarizing your saved chat messages from a local SQLite database. Requires a companion setup in the ./chatbot directory to collect/store chat logs. Configured via CHAT_DB_PATH env var and typically run via Node/Claude Desktop stdio integration. Provides the tool `query_chat_messages` to query chat messages with parameters and generate summaries based on a prompt.
Stars: 1.0k
Forks: 100
License: Not specified
Category: Specialized
Overview
Installation
CHAT_DB_PATH=path-to/chatbot/data/chat.db pnpm install pnpm build pnpm watch{
"mcpServers": {
"mcp-server-chatsum": {
"command": "path-to/bin/node",
"args": ["path-to/mcp-server-chatsum/build/index.js"],
"env": {
"CHAT_DB_PATH": "path-to/mcp-server-chatsum/chatbot/data/chat.db"
}
}
}
}pnpm inspectorquery_chat_messages
Query chat messages with given parameters and summarize chat messages based on the query prompt
FAQs
Why does ChatSum MCP Server require two separate processes, and what happens if the chatbot goes offline?
ChatSum separates concerns: the chatbot handles WeChat automation and real-time message capture into SQLite, while the MCP server provides stateless query access. If the chatbot goes offline, no new messages are captured, creating permanent data gaps since WeChat disallows retroactive imports. The MCP server continues serving historical queries from existing records but cannot retrieve missed conversations.
How do I configure ChatSum MCP Server to work with MCP clients other than Claude Desktop, such as Cursor or VS Code?
Adapt Claude Desktop's JSON configuration to your client's MCP settings: provide absolute paths to your Node binary in the command field and ChatSum's build/index.js in args, then set CHAT_DB_PATH as an environment variable pointing to your SQLite database. Cursor and VS Code use similar JSON-based MCP configurations, though exact locations vary.
Is there an actively maintained MCP server specifically for querying WeChat conversations?
No. ChatSum is the only known MCP server built specifically for WeChat, but it has been inactive since December 2024 with no maintainer responses to open issues. While the core functionality exists, production use requires accepting account suspension risk from outdated WeChat automation libraries and maintaining a continuously-running companion chatbot process without upstream support.
What are the best alternatives to ChatSum MCP Server for accessing chat messages through MCP?
The Slack MCP server offers OAuth authentication with message, thread, and canvas access. Telegram's overpod includes twenty tools with hosted options. macOS users can access iMessage's native database with date filtering. The rebots-online server enables semantic search via vector embeddings, allowing natural language queries across multiple platforms simultaneously instead of platform-specific lookups.
What are the risks of using the wechaty-puppet-wechat4u library for WeChat automation, and can my account get banned?
Yes, WeChat accounts can be banned using wechaty-puppet-wechat4u because WeChat actively detects and blocks RPA tools. The library simulates the web protocol, considered unauthorized access. Bans can be temporary or permanent, triggered by unusual login patterns, message volume, or automation signatures. Being unmaintained means no security patches or evasion updates, increasing ban risk.
Can I use MCP servers with AI tools other than Claude Desktop?
Yes. MCP is an open protocol, so any client implementing the standard can connect to MCP servers via JSON-RPC handshakes. Tools like Cursor, Windsurf, Jan AI, Azure AI Agents, and Amazon Quick Agents support MCP integration, though specific server compatibility depends on transport type like stdio or SSE and individual client implementations.