api-to-curl-mcp-server

🚀 MCP-AI: Self-Learning API-to-cURL Model

This project builds an autonomous AI system to convert API documentation into cURL commands.

📌 Features:

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

---

🚀 Quick Start:

1️⃣ Install dependencies:

pip install -r requirements.txt

2️⃣ Start MCP Server:

bash scripts/start_mcp.sh

3️⃣ Start AI Automation:

python src/ai_autonomous_dev.py

4️⃣ Test System:

pytest tests/

---

📜 setup.py (For Packaging SDK)

from setuptools import setup, find_packages

setup(
    name="mcp_sdk",
    version="1.0",
    packages=find_packages(),
    install_requires=[
        "fastapi",
        "uvicorn",
        "torch",
        "transformers",
        "sacrebleu",
        "requests",
        "pytest",
        "gitpython",
    ],
    author="Your Name",
    description="MCP SDK for API-to-cURL Model Automation",
    license="MIT"
)

---

✅ Final Steps

1️⃣ Install dependencies

pip install -r requirements.txt

2️⃣ Start MCP Server

bash scripts/start_mcp.sh

3️⃣ Run AI Automation

python src/ai_autonomous_dev.py

4️⃣ Test System

pytest tests/

Fix uvicorn: command not found

The error indicates that uvicorn is not installed or not in the system path.

✅ Solution 1: Install Uvicorn

pip install uvicorn

✅ Solution 2: Ensure Virtual Environment is Activated

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

✅ Solution 3: Explicitly Call Python for Uvicorn

Modify scripts/start_mcp.sh to:

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

Final Steps

After applying the fixes, restart everything:

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

🚀 Now the system is fully organized and self-learning! 🎯