CowAgent
CowAgent (chatgpt-on-wechat) 是基于大模型的超级AI助理，能主动思考和任务规划、访问操作系统和外部资源、创造和执行Skills、通过长期记忆和知识库不断成长，比OpenClaw更轻量和便捷。同时支持微信、飞书、钉钉、企微、QQ、公众号、网页等接入，可选择DeepSeek/OpenAI/Claude/Gemini/ MiniMax/Qwen/GLM/LinkAI，能处理文本、语音、图片和文件，可快速搭建个人AI助理和企业数字员工。
What is CowAgent?
CowAgent is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to cowagent (chatgpt-on-wechat) 是基于大模型的超级ai助理，能主动思考和任务规划、访问操作系统和外部资源、创造和执行skills、通过长期记忆和知识库不断成长，比openclaw更轻量和便捷。同时支持微信、飞书、钉钉、企微、qq、公众号、网页等接入，可选择deepseek/openai/claude/gemini/ minimax/qwen/glm/linkai，能处理文...
CowAgent (chatgpt-on-wechat) 是基于大模型的超级AI助理，能主动思考和任务规划、访问操作系统和外部资源、创造和执行Skills、通过长期记忆和知识库不断成长，比OpenClaw更轻量和便捷。同时支持微信、飞书、钉钉、企微、QQ、公众号、网页等接入，可选择DeepSeek/OpenAI/Claude/Gemini/ MiniMax/Qwen/GLM/LinkAI，能处理文本、语音、图片和文件，可快速搭建个人AI助理和企业数字员工。
This server falls under the Business Applications category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- CowAgent (chatgpt-on-wechat) 是基于大模型的超级AI助理，能主动思考和任务规划、访问操作系统
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx cowagentConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use CowAgent
CowAgent (formerly chatgpt-on-wechat) is an open-source AI agent harness that connects large language models to 11+ messaging platforms — including WeChat, Telegram, Slack, Discord, Feishu, DingTalk, and WeCom — while supporting autonomous task planning, long-term memory, personal knowledge bases, and self-evolving skill creation. It integrates with DeepSeek, Claude, OpenAI, Gemini, Qwen, GLM, Kimi, and other LLMs through a unified config, and supports MCP tools via a single mcp.json file with stdio and SSE transports. Teams and individuals use CowAgent to build personal AI assistants or enterprise digital employees that can browse the web, run terminal commands, process files, and install new skills on demand.
Prerequisites
- Linux, macOS, or Windows (or Docker) for running the CowAgent server
- An API key for at least one supported LLM: Claude (ANTHROPIC_API_KEY), OpenAI (OPENAI_API_KEY), DeepSeek, or similar
- At least one messaging platform account/bot token (e.g. Telegram bot token, Slack app credentials)
- Docker and Docker Compose (recommended for easiest deployment)
- Port 9899 available for the CowAgent web interface
Install CowAgent
Run the one-line installer for your platform. This downloads CowAgent and sets up the web interface at localhost:9899.
# Linux/macOS:
bash <(curl -fsSL https://cdn.link-ai.tech/code/cow/run.sh)
# Windows PowerShell:
irm https://cdn.link-ai.tech/code/cow/run.ps1 | iex
# Docker:
curl -O https://cdn.link-ai.tech/code/cow/docker-compose.yml
docker compose up -dConfigure your LLM provider
Open the CowAgent web interface at http://localhost:9899 and configure your LLM provider, API key, and model selection. Or edit config.json directly.
# Key config.json settings:
# {
# "web_host": "0.0.0.0",
# "web_password": "your-secure-password",
# "web_port": 9899,
# "model": "claude-3-5-sonnet-20241022",
# "open_ai_api_key": "sk-ant-..."
# }Connect a messaging platform
Through the web interface or config.json, connect your preferred messaging channel. For Telegram, provide your bot token. For Slack, provide your app's OAuth token and signing secret.
Configure MCP tools
Create an mcp.json file to enable MCP tools for CowAgent. The file supports both stdio and SSE transports and hot-reloads when changed.
# mcp.json example:
{
"mcpServers": {
"filesystem": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-filesystem", "/path/to/allowed/dir"]
}
}
}Install skills from the marketplace
Extend CowAgent's capabilities by installing community skills through the web interface or chat commands. Skills add new autonomous capabilities like report generation or data processing.
/skill install web-search
/skill install code-runner
/skill listCowAgent Examples
Client configuration
CowAgent itself acts as an AI assistant hub. To connect it as an MCP server to Claude Desktop, point the config at your running CowAgent instance.
{
"mcpServers": {
"cowagent": {
"command": "npx",
"args": ["cowagent"],
"env": {
"COWAGENT_URL": "http://localhost:9899",
"COWAGENT_PASSWORD": "your-web-password"
}
}
}
}Prompts to try
Use these prompts through any connected messaging platform or the web interface to explore CowAgent's autonomous capabilities.
- "Search the web for the latest news about AI agents and send me a daily summary at 9am"
- "Create a skill to automatically generate weekly project status reports from my task list"
- "Search my knowledge base and summarize everything we know about our API authentication system"
- "Schedule a daily memory consolidation task to organize my conversation history"
- "Read the file at /projects/report.csv and create a chart showing monthly trends"Troubleshooting CowAgent
Web interface at localhost:9899 is unreachable after installation
Check if the CowAgent process started successfully by reviewing logs. For Docker, run 'docker compose logs cow'. For native installs, ensure port 9899 is not already in use ('lsof -i :9899'). If running on a remote server, set web_host to '0.0.0.0' in config.json.
Messages sent to Telegram/WeChat receive no response
Verify the bot token or platform credentials are correctly configured. For Telegram, confirm your bot is not blocked and the webhook or polling mode is running. Check CowAgent logs for API call errors — often caused by incorrect API keys or model names.
MCP tools are not being invoked even after configuring mcp.json
CowAgent hot-reloads mcp.json but the tool names must match what your LLM model expects to call. Ensure the mcp.json is valid JSON and the server commands are executable. Restart CowAgent after significant config changes to force a full reload.
Frequently Asked Questions about CowAgent
What is CowAgent?
CowAgent is a Model Context Protocol (MCP) server that cowagent (chatgpt-on-wechat) 是基于大模型的超级ai助理，能主动思考和任务规划、访问操作系统和外部资源、创造和执行skills、通过长期记忆和知识库不断成长，比openclaw更轻量和便捷。同时支持微信、飞书、钉钉、企微、qq、公众号、网页等接入，可选择deepseek/openai/claude/gemini/ minimax/qwen/glm/linkai，能处理文本、语音、图片和文件，可快速搭建个人ai助理和企业数字员工。 It connects AI assistants to external tools and data sources through a standardized interface.
How do I install CowAgent?
Follow the installation instructions on the CowAgent GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with CowAgent?
CowAgent works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is CowAgent free to use?
Yes, CowAgent is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
CowAgent Alternatives — Similar Business Applications Servers
Looking for alternatives to CowAgent? Here are other popular business applications servers you can use with Claude, Cursor, and VS Code.
n8n
★ 189.1kA comprehensive MCP server that provides full control over n8n automation workflows through natural language. It offers 43 tools for managing workflows, executions, credentials, and data tables, with safety features like write-mode protection and dou
LobeHub
★ 77.5k🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.
Jeecgboot
★ 46.4kAI 低代码平台，「低代码 + 零代码」双模式驱动：低代码一键生成前后端代码，零代码 5 分钟搭建系统，AI Skills 一句话画流程、设计表单、生成整套系统。内置 AI聊天、知识库、流程编排、MCP插件等，兼容主流大模型。引领「AI 生成 → 在线配置 → 代码生成 → 手工合并->AI修改」开发模式，消除 Java 项目 80% 的重复工作，提效而不失灵活。
Minds Platform
★ 39.2kPlatform dedicated to building an open foundation for applied Artificial Intelligence, designed for people seeking production-ready AI systems they can truly control, extend and deploy anywhere.
Astrbot
★ 32.8kAI Agent Assistant & development framework that integrates lots of IM platforms, LLMs, plugins and AI feature, and can be your openclaw alternative. ✨
FastGPT
★ 28.1kFastGPT is a knowledge-based platform built on the LLMs, offers a comprehensive suite of out-of-the-box capabilities such as data processing, RAG retrieval, and visual AI workflow orchestration, letting you easily develop and deploy complex question-
Browse More Business Applications MCP Servers
Explore all business applications servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up CowAgent in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use CowAgent?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.