🛡️ GenAI Security Agent – Capstone 2025

A Generative AI-powered assistant that detects, interprets, and provides real-time remediation recommendations for cloud security misconfigurations using function calling, severity scoring, and simulated LangGraph workflows.

🔍 What This Project Demonstrates

• Prompt engineering with Gemini and simulated LangGraph agents
• Misconfiguration detection in IAM policies and firewall configs
• Severity scoring and risk classification
• Real-time grounding via SerpAPI for dynamic remediation tips
• Automated reports in Markdown, PDF, and HTML formats

📁 Repository Structure

/
├── Assets/
│   ├── genai_security_logo.png
│   └── risk_breakdown_chart.png
├── Notebook/
│   └── genai-security-agent-capstone-2025-yuwan.ipynb
├── Reports/
│   ├── Security_Misconfig_Report.md
│   ├── Security_Misconfig_Report.pdf
│   └── Security_Misconfig_Report.html

🧪 Technologies Used

• Gemini Pro (simulated)
• Python (LangGraph simulation)
• SerpAPI (real-time search grounding)
• Matplotlib, Counter, FPDF
• Markdown, HTML, PDF exports

🪄 Enhancements (Future Work)

• Integrate with live IAM and GCP log data
• Connect to dashboards (ELK, Splunk)
• Full LangGraph integration

🧑‍💻 Creator

Created with purpose by Yuwanthiran Sukalingam (Yuwan)
🎯 Vision: Secure AI-powered infrastructure for DevSecOps, SecOps, and Compliance.

📸 Logo Preview

📊 Risk Breakdown Sample

🔖 Submission Hashtags

#genai-security, #cloud-misconfig-detection, #kaggle-ai-capstone, #langgraph, #gemini, #secops, #yuwan