MCP-AI Server
by S-Umasankar
This project builds an autonomous AI system to convert API documentation into cURL commands. It features a self-improving model and an MCP server for API-based execution.
Last updated: N/A
What is MCP-AI Server?
The MCP-AI server is an autonomous AI system designed to automatically convert API documentation into cURL commands. It utilizes a self-improving model based on reinforcement learning and provides an API-based execution environment.
How to use MCP-AI Server?
To use the MCP-AI server, first install the dependencies using pip install -r requirements.txt. Then, start the MCP server with bash scripts/start_mcp.sh and initiate the AI automation by running python src/ai_autonomous_dev.py. Finally, test the system using pytest tests/.
Key features of MCP-AI Server
Automated Dataset Generation
Self-Improving Model with Reinforcement Learning
MCP Server for API-based Execution
Continuous Deployment with GitHub Actions
Use cases of MCP-AI Server
Automated API testing
Rapid cURL command generation from API documentation
Simplifying API interaction for developers
Self-learning API integration
Streamlining API workflow
FAQ from MCP-AI Server
How do I fix 'uvicorn: command not found'?
How do I fix 'uvicorn: command not found'?
Install Uvicorn using pip install uvicorn, ensure your virtual environment is activated, or explicitly call Python for Uvicorn in the scripts/start_mcp.sh file.
What are the dependencies for this project?
What are the dependencies for this project?
The dependencies include fastapi, uvicorn, torch, transformers, sacrebleu, requests, pytest, and gitpython, as specified in the setup.py file.
How do I start the MCP server?
How do I start the MCP server?
Run the script bash scripts/start_mcp.sh.
How do I run the AI automation?
How do I run the AI automation?
Execute the command python src/ai_autonomous_dev.py.
How do I test the system?
How do I test the system?
Run the command pytest tests/.