Lab Virtual MCP Server

execution in a secure and scalable lab setup.

⚙️ Lab Virtual MCP Server (Execute Code Remotely via Claude AI) Create a virtual lab for users to run custom code remotely using the Model Context Protocol (MCP) and integrate with Claude AI or other clients.

🔗 GitHub Repo

📂 This repo has moved to base/base-mcp

🚀 What This Lab Server Does 🧠 Enables remote Python code execution through cline AI

🧪 Supports real-time lab scenarios (code evaluation, sandbox testing, etc.)

📋 Prerequisites Ensure you have the following on your system:

✅ Python 3.10.11

✅ pip (Python package manager)

✅ fastmcp (to serve the MCP endpoint)

✅ uv (virtual environment manager, via scoop or curl)

✅ Access to Claude Desktop or Cursor or cline (for testing)

🧱 Installation Steps

1. Clone the MCP Server Repo

2. Set up Python Environment

cd McpSever_Remote_code_execution

3. Set Up Virtual Environment

4. Install Dependencies

5. Run the Server

You now have a remote code execution server listening for requests via MCP.

🧪 MCP Client Configuration For Claude Desktop / Cursor, update your mcp_config.json:

Beofre start Mcp set path

✅ Available Tools (Prebuilt in MCP)

Tool Description execute_code Executes user-provided Python code

💡 Recommendations for Lab Admins ✅ Add sandboxing logic to app.py if users can run arbitrary code.

✅ Use Docker or subprocess isolation for safer execution (optional).

✅ Monitor logs and set execution timeouts.