Winner of Google Developer Groups @ McMaster Hackathon 🏆

CodeTurret is an automated security architect designed for the era of vibe-coding. Point it at any GitHub repository and it will scan for vulnerabilities, stream results live, and automatically open a pull request with AI-generated fixes.

• Dual-Pass AI Scanning -> Gemini Flash for rapid triage across all files, Gemini Pro for deep analysis on high-severity findings
• Real-Time Streaming -> scan progress streams file-by-file via Server-Sent Events; no waiting for a blocking response
• Auto-Fix PRs -> after a scan, one click generates AI patches for every vulnerability and opens a GitHub Pull Request
• Git Intelligence -> git blame, hot-file detection, and security commit analysis enrich every finding with authorship context
• Ask Cortex -> natural-language security consultant powered by Snowflake Cortex AI; ask "who introduced the SQL injection?" and get an answer
• Async Job Queue -> scans are processed via RabbitMQ workers; the API returns instantly and the UI streams progress

Next.js Frontend
      │
      ├── POST /api/scan  ─────────────────►  RabbitMQ [scan.requests]
      ├── GET  /api/scans/{id}/stream  ◄────  SSE ← RabbitMQ [scan.progress]
      ├── GET  /api/findings/{scanId}
      └── POST /api/ask  ──────────────────►  Snowflake Cortex AI

Spring Boot 3.3 (Java 21)
      ├── ScanWorker      — clones repo, runs Gemini Flash + Pro, persists findings
      └── FixWorker       — generates patches with Gemini Pro, pushes branch, opens PR

Storage
      ├── PostgreSQL      — repos, scans, findings, fix PRs
      └── Snowflake       — Cortex AI for natural-language Q&A over findings

• Java 21
• Maven
• Docker Desktop (for PostgreSQL + RabbitMQ)
• A Gemini API key
• Snowflake account (for the Ask feature — optional)

# 1. Start PostgreSQL + RabbitMQ
cd backend
docker-compose up -d

# 2. Configure environment
cp .env.example .env
# Fill in GEMINI_API_KEY, ENCRYPTION_SECRET_KEY
# Add SNOWFLAKE_* credentials if you want the Ask feature

# 3. Start the backend (Flyway auto-creates tables on first run)
mvn spring-boot:run

# 4. Start the frontend
cd ../frontend
cp .env.local.example .env.local
npm install && npm run dev

Backend: http://localhost:8080 · Frontend: http://localhost:3000

Homepage	Scanner	Reports