Natural Language Shell

AI-powered terminal — type plain English, execute Unix commands. Custom C shell + Gemini AI translation with real-time web UI.

---

What It Does

Converts natural language like "find all python files modified today" into the correct shell command, executes it through a custom C-based shell, and streams results in real time via WebSockets.

Technical Highlights:

• Custom shell in C — fork/exec/wait, piping, redirection, job control
• AI translation — Google Gemini converts English → Unix commands (30+ patterns)
• Dual execution layer — custom shell + native fallback for unsupported commands
• Real-time streaming — REST + WebSocket, end-to-end latency < 100ms
• System-wide file search — directory traversal with path resolution
• Voice input — Web Speech API for hands-free operation

Architecture

React (Terminal UI + Voice) → Flask API → Gemini AI (command translation)
                                              → Custom C Shell (fork/exec)
                                              → Fallback: Native System Shell
                              ← WebSocket (live output stream)

Tech Stack

Component	Technology
Shell	C (process mgmt, pipes, redirection)
Backend	Python, Flask, Google Gemini API
Frontend	React + WebSocket
Voice	Web Speech API

My Role

I designed the dual-execution architecture, built the C shell's process management model, planned the command translation pipeline, and chose WebSockets for real-time output. Code generation was accelerated using AI tools; systems architecture and process lifecycle management are mine.

Quick Start

git clone https://github.com/AdityaPandey-DEV/Natural-Language-Shell.git && cd Natural-Language-Shell
make                              # Build C shell
pip install -r requirements.txt   # Python deps
python shell_bridge.py            # Start backend
cd frontend && npm install && npm run dev   # Start UI

---

Architected & built by Aditya Pandey — AI-augmented development