MCP Agent Server

Description

This project appears to be a server and client implementation for the Model Context Protocol (MCP), designed to facilitate interactions with AI agents. It includes components for a command-line interface (CLI), a web-based interface, and a dedicated module for code execution.

Features

• Command-Line Interface (CLI): Allows interaction with MCP agents via agentcli.py.
• Web Interface: Provides a web-based UI for interacting with MCP agents, served by agentweb.py and located in the frontend/ directory.
• MCP Code Executor: A Node.js module (mcp_code_executor/) likely responsible for executing code snippets or commands as directed by agents.
• Local MCP Modules: Python modules (mcp_local_modules/) for handling MCP setup, configuration, and utility functions.
• Documentation: Contains project-related documentation in the docs/ directory.
• Sample MCP Files: Includes example files related to MCP in sample_mcp_files/.

Project Structure

A brief overview of the key directories and files:

• agentcli.py: Main script for the command-line interface.
• agentweb.py: Main script for the web server.
• frontend/: Contains the HTML, CSS, and JavaScript for the web interface.
• mcp_code_executor/: A Node.js module for code execution.
  • src/index.ts: TypeScript source for the code executor.
  • build/index.js: Compiled JavaScript for the code executor.
• mcp_local_modules/: Python modules for MCP core functionalities.
  • mcp_agent_setup.py: Handles agent setup.
  • mcp_server_config.py: Manages server configuration.
• cli/: Python modules supporting the command-line interface.
• docs/: Project documentation files.
• requirements.txt: Python project dependencies.
• .gitignore: Specifies intentionally untracked files that Git should ignore.
• file_organization_log.md: Log related to file organization.
• minimal_ollama_test.py: Test script for Ollama.
• ollama_chat.py: Script for Ollama chat functionality.

Getting Started

To get started with this project, you would typically need to:

1. Install Dependencies:
   • For Python components: pip install -r requirements.txt
   • For the MCP Code Executor: Navigate to mcp_code_executor/ and run npm install
2. Configure the Environment: (Details would depend on specific MCP server configurations needed)
3. Run the Application:
   • CLI: python agentcli.py (or similar, depending on arguments)
   • Web Interface: python agentweb.py (and then access the specified port in a browser)

(Please update with more specific instructions as the project develops.)

Contributing

Contributions are welcome! Please refer to contribution guidelines if available, or open an issue/pull request.

License

(Please specify the license under which this project is distributed.)