🤖 AI Debug Agent — Ultra (Groq Edition)

100% FREE | Powered by Groq API | Real-time streaming | Llama 3.3 70B | ZIP/TAR support

---

🚀 Setup (3 minutes)

Step 1 — Install Node.js

Download from nodejs.org — v18 or higher needed

Step 2 — Get FREE Groq API Key

1. Go to console.groq.com
2. Sign up (free, no credit card)
3. Click API Keys → Create API Key
4. Copy the key (starts with gsk_...)

Step 3 — Configure

# Copy env file
cp .env.example .env

# Open .env and paste your key:
GROQ_API_KEY=gsk_your_actual_key_here

Step 4 — Install & Start

npm install
npm start

Step 5 — Open Browser

http://localhost:3000

---

🦙 Free Groq Models

Model	Speed	Smarts	Best For
llama-3.3-70b-versatile ⭐	Fast	🔥🔥🔥	All-round debugging (default)
deepseek-r1-distill-llama-70b	Medium	🧠🧠🧠	Complex logic bugs
llama-3.1-8b-instant	Ultra fast	🔥🔥	Quick checks
mixtral-8x7b-32768	Fast	🔥🔥🔥	Large files (32K context)
gemma2-9b-it	Fast	🔥🔥	Lightweight use

Change model in .env:

GROQ_MODEL=deepseek-r1-distill-llama-70b

---

⚡ Features

• 📦 ZIP, TAR.GZ, TAR, GZ — extract all code files automatically
• 🤖 Groq AI Debugging — Llama 3.3 70B finds every bug type
• ⚡ Real-time Streaming — see analysis live via WebSocket
• 🔀 Split Diff Viewer — side-by-side original vs fixed
• 🔁 Batch Processing — 3 files at once (configurable)
• ⏸ Pause / Resume — mid-batch control
• 📅 Auto Scheduling — run every X minutes/hours
• ⬇ Export ZIP — all fixed files + markdown report
• 🛡 Security Scanning — SQLi, XSS, path traversal, etc.
• 30+ Languages supported

---

⌨️ Keyboard Shortcuts

Key	Action
Ctrl+O	Upload files
Ctrl+Enter	Run debug batch
Ctrl+D	Download fix
Ctrl+L	Clear terminal

---

📁 Files

ai-debug-agent/
├── server.js          ← Node.js backend (Groq API + WebSocket)
├── public/index.html  ← Full frontend UI
├── package.json
├── .env               ← Your config (create from .env.example)
└── README.md

---

❓ Troubleshooting

"GROQ_API_KEY not configured" → Create .env from .env.example, paste your key from console.groq.com

Rate limit errors → Groq free tier has limits. Lower BATCH_CONCURRENCY=1 in .env

Large file issues → Use mixtral-8x7b-32768 model (supports 32K tokens)