api-to-curl-mcp-server

MCP-AI is a cutting-edge project that leverages an autonomous AI system to transform API documentation into cURL commands. The system is designed to be self-improving through reinforcement learning, ensuring that it becomes more efficient over time. It features an MCP server for API-based execution and supports continuous deployment using GitHub Actions. The project is structured to facilitate easy setup and execution, with scripts provided for installing dependencies, starting the MCP server, running AI automation, and testing the system. The packaging of the SDK is managed through a setup script, ensuring that all necessary components are included for seamless operation.

Features

• Automated Dataset Generation
• Self-Improving Model with Reinforcement Learning
• MCP Server for API-based Execution
• Continuous Deployment with GitHub Actions

Usage with Different Platforms

Install dependencies

bash
pip install -r requirements.txt

Start MCP Server

bash
bash scripts/start_mcp.sh

Run AI Automation

bash
python src/ai_autonomous_dev.py

Test System

bash
pytest tests/

Fix uvicorn: command not found

bash
pip install uvicorn

Ensure Virtual Environment is Activated

bash
source /Users/umasankars/PycharmProjects/CapstoneMCPserver/venv/bin/activate
pip install -r requirements.txt

Explicitly Call Python for Uvicorn

bash
#!/bin/bash
echo "🚀 Starting MCP Server..."
/Users/umasankars/PycharmProjects/CapstoneMCPserver/venv/bin/python -m uvicorn src.mcp_server:app --reload

Final Steps

bash
pip install --upgrade pip setuptools wheel
pip install -r requirements.txt
bash scripts/start_mcp.sh