M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
W

WeChat MCP Server

Community
MCP server that lets AI assistants read and reply to WeChat
How to install ↓ Compare GitHub source repository ↗
Category
Other #158 of 230
Stars
★ 276 Popular
Transport
stdio (local process) · SSE · Streamable HTTP
Runtime
Python 3.12+
Credentials
No credential needed
License
MIT
Last commit
⚠
Tools
4
49FMRS · D

WeChat MCP Server is a feature-rich MCP server designed for macOS users, leveraging the Accessibility API to control WeChat. It offers practical tools for fetching messages, replying, adding contacts, and publishing moments, and includes 5 Claude Code sub-agents to enhance automation. Installation is straightforward but requires careful permission configuration.

Strongest · Setup experience 13/20 Weakest · Reliability 8/20

Reliability
8/20
Security and permissions
8/20
Maintenance
8/20
Documentation
12/20
Setup experience
13/20
Why each score
Reliability 8/20
Supplied evidence is README + metadata only; no source, tests, CI, or runnable artifacts were provided. The README describes a plausible stdio/streamable-http/SSE server with four tools and Python 3.12+, but tool/behavior parity and error handling cannot be verified. Deductions: no committed tests or CI workflows, no failure-mode documentation, and platform coupling to macOS Accessibility API and screen capture makes clean startup uncertain. Score is kept below the 12 static cap because no execution evidence exists.
Security and permissions 8/20
No red-line issues are visible in the supplied material. The main risks are explicit: Accessibility API and screen capture grant broad read access to WeChat/screen content, and tools can send replies, add contacts, and publish Moments. The README mentions optional draft-only Moments and configurable privacy for contact add, but there is no confirmation mechanism, least-privilege scoping, credential/data-flow disclosure, or external-network behavior described. Score reflects visible risks with incomplete permission/confirmation controls; no source is available to inspect deeper.
Maintenance 8/20
The repository is not archived and has an MIT license, and the README references a PyPI package and documentation, which suggests an existing project structure. However, supplied metadata contains no commit cadence, release history, issue-response evidence, dependency-update policy, or security-response channel; the 10 open issues cannot be assessed. Stars are used only as a discovery signal, not for scoring. Deductions are made for missing maintenance governance evidence.
Documentation 12/20
The README gives install commands, client configuration examples (Claude Code, Claude Desktop, Codex), transport options, tool names, macOS permission steps, and links to a detailed guide and sub-agent docs, so it is usable as a first-layer document. Deductions: no troubleshooting section, no explicit limitations beyond macOS/WeChat, no error/cost/limits details, and the linked detailed docs were not supplied for verification, so hidden assumptions remain.
Setup experience 13/20
Setup is documented as a few steps: pip install, register the stdio command in a mainstream client, grant Accessibility permission, and run WeChat. Config examples are given for three clients, which is good. Deductions: it is macOS-only and depends on a running WeChat and manually granted OS permission, with no automated dependency checks or verified end-to-end flow; no CI/tests were supplied, so the setup score is capped below 15 under static calibration.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessUses the networkControls a browser

Best for

  • macOS users
  • Developers using Claude Code or Claude Desktop
  • Individuals or teams needing WeChat automation integration
  • Users comfortable with command line and Python environments

Not for

  • Non-macOS platforms (Windows, Linux)
  • Users without WeChat for Mac or accessibility permissions
  • Publishing Moments with media (photos, videos)
  • Users needing cross-device or mobile support

Required permissions

  • macOS Accessibility permissions (to control WeChat)
  • WeChat for Mac installed and running

Risks and side effects

  • Relies on macOS Accessibility API, which may be affected by system updates
  • Granting accessibility permissions to terminal poses security risks
  • Changes to WeChat's interface may cause tools to fail
  • Automated actions may violate WeChat's terms of service
  • Accessibility-based control may be detected by WeChat as automation

Setup

Before you start

Runtime：Python 3.12+

  1. Install: pip install wechat-mcp-server
  2. For Claude Code: claude mcp add --transport stdio wechat-mcp -- wechat-mcp (if installed via pip) or claude mcp add --transport stdio wechat-mcp -- uv --directory $(pwd) run wechat-mcp (if using uv for development).
  3. Alternatively, configure Claude Desktop with the provided JSON example.
  4. For Codex: codex mcp add wechat-mcp -- wechat-mcp.
  5. Grant Accessibility permissions to your terminal: System Settings → Privacy & Security → Accessibility, add your terminal app.
  6. Ensure WeChat for Mac is installed and running.

Check that it works

After granting Accessibility permissions to your terminal and starting wechat-mcp, the client's tool list should show fetch_messages_by_chat and the other tools; asking the AI to fetch recent messages from any chat confirms the connection.

Troubleshooting

  1. Ensure accessibility permissions are granted (System Settings → Privacy & Security → Accessibility)
  2. Ensure WeChat for Mac is installed and running
  3. Check Python version meets requirements (3.12+)
  4. If using uv, ensure `uv sync` has been executed
  5. Try different transport methods (stdio, sse, streamable-http)

Things to try

Once connected, you can ask your AI assistant things like:

  • Fetch my recent messages from a chat with one of my contacts
  • Based on the chat history, draft and send a reply
  • Add a new contact by WeChat ID and send a friend request
  • Publish a text-only Moments post, but just prepare it as a draft without posting

Tools 4

fetch_messages_by_chat read-only
Get recent messages from a chat
reply_to_messages_by_chat writes
Send a reply to a chat
add_contact_by_wechat_id writes
Add a new contact using a WeChat ID and send a friend request
publish_moment_without_media writes
Publish a text-only Moments post (no photos or videos); optionally only prepare a draft without posting via publish=False

Use cases

Automate replies to WeChat messages
Monitor chats and summarize information
Search chat history intelligently
Publish Moments drafts
Manage contacts via AI

Supported clients

Claude Desktop
Claude Code
Codex
OpenClawPartial support
ChatGPTPartial support

Listed from the project's documentation, not tested by this site.

Overview

WeChat MCP Server is an MCP server that automates WeChat on macOS using the Accessibility API and screen capture. It enables LLMs to interact with WeChat chats programmatically. Features include fetching recent messages from any chat (contact or group), sending automated replies, publishing text-only Moments posts (with optional draft-only mode), adding contacts using WeChat ID with configurable privacy, smart chat search with exact name matching, and 5 specialized Claude Code sub-agents for smart WeChat automation.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 654 · Tools 11 Compare with this →
Compare with other MCP servers

Source revision cc470367ca66 Data synced 2026-10-12 Read the FMRS scoring method