mcpField
M
MCP Agent

MCP Agent

github.com/justinhuang0208Desktop Apps2Updated 37w ago
Repository

About MCP Agent

A LaunchBar desktop application integrated with Langchain and MCP to automatically complete tasks based on prompts. It dynamically loads MCP servers and provides autocomplete suggestions for available tools.

Key features

  1. 01Langchain and MCP integration
  2. 02Dynamic MCP server loading
  3. 03Autocomplete with @ symbol
  4. 04LLM API key configuration
  5. 05Tool-based server selection

Use cases

  1. 01Quick command execution in LaunchBar
  2. 02Automated multi-MCP service management
  3. 03Cross-service task automation

MCP support

The client connects to and manages MCP servers via configuration file, dynamically loading the appropriate MCP server based on the tool name specified.

Details

From the project README

MCP Agent

This is a Launchbar action integrated with Langchain and MCP to automatically complete tasks based on your prompt.

Configuration

Before using the application, ensure that you add a .env file to configure your LLM API key. Additionally, include an mcp_config.json file to set up your MCP.

The mcp_config.json should format like this:

{
  "filesystem": {
    "command": "npx",
    "args": [
      "-y",
      "@modelcontextprotocol/server-filesystem",
      "/your/folder/path1",
      "/your/folder/path2"
    ]
  },
  "fetch": {
    "command": "uvx",
    "args": [
      "mcp-server-fetch"
    ]
  }
}

The suggestions.js file is designed to activate autocomplete when you type "@", providing suggestions for the MCP server names. Furthermore, this function dynamically loads only the MCP server associated with the tool name you've specified. You have the flexibility to modify the keyword section within suggestions.js to align with the MCP server you have installed or to alter its name if you want.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.