An AI-Powered Security Scanner with Automated Defense Strategies
MINA (Minimalistic Intelligent Network Agent) is a lightweight pentesting assistant that combines rule-based vulnerability detection with AI-powered analysis and defensive recommendations.
- Hybrid scanning: Rule-based + AI verification
- Detects: SQL Injection (SQLi), XSS, Path Traversal
- Configurable timeout & AI toggle
- AI-generated mitigation strategies
- Practical, implementable remediation steps
- Contextual vulnerability explanations
- Interactive CLI menu
- Traditional command-line mode
- Rich terminal output with progress tracking
- Auto-generated JSON reports
- Scan & mitigation logs
- Color-coded vulnerability summaries
-
Security Students & Researchers → Learn basics of vulnerability detection (SQLi, XSS, Path Traversal).
-
Developers → Quickly scan personal apps before deploying.
-
Consultants → Use as a lightweight pentesting assistant for demos.
-
Bug Bounty Hunters → Rule-based + AI-powered anomaly checks for fast triage.
-
Language: Python 3.10+
-
Libraries:
requests → HTTP requests & payload injection BeautifulSoup → Response parsing tenacity → Retry handling for stable scans json → Report generation rich → Beautiful CLI output AI/LLM : Ollama (Phi model) for anomaly detection & mitigation generation
git clone https://github.com/yokodrea/Mina.git
cd Mina
pip install -r requirements.txt
python main.py
🧩 Modular Plugin System → Add custom vulnerability tests.
🌐 Expanded Vulnerability Coverage → CSRF, SSRF, Command Injection, etc.
🤖 Multi-Model AI Analysis → Integrate GPT, Llama, or custom fine-tuned models.
📊 Dashboard → Rich visualization of scan results & risk scoring.
☁️ Cloud Mode → Centralized scanning with API support for teams.
🔑 Authentication Support → Handle login-protected endpoints.
📌 To be filled in once final dependencies are set. 👉 See requirements.txt for more info.
This project is licensed under the MIT License.