PolicyForge — AI Agent Security Policy Platform

TechEx Intelligent Enterprise Solutions Hackathon 2026 Track 1: Agent Security & AI Governance | Prize Pool: $10,000

PolicyForge lets enterprise security teams write AI agent security policies in plain English. Gemini 2.0 Flash converts them to Lobster Trap YAML enforcement rules — no YAML knowledge needed.

---

Demo Flow (2 minutes)

1. Policy Editor → Type: "Block any agent that tries to read patient SSN or medical records"
2. Generate → Gemini creates a Lobster Trap YAML rule in ~2 seconds
3. Activate → Policy enforced immediately
4. Attack Demo → Fire "PII Exfiltration" attack → See BLOCK with 95% risk score
5. Dashboard → Blocked count increments, audit trail updates live
6. Compliance Report → Generate HIPAA report → Download .md

---

Architecture

┌─────────────────────────────────────┐
│   PolicyForge UI (Next.js 15)       │
│   Dashboard | Editor | Demo | Report│
└──────────────┬──────────────────────┘
               │ REST API
┌──────────────▼──────────────────────┐
│   PolicyForge API (FastAPI)         │
│   Gemini 2.0 Flash | SQLite         │
└──────────────┬──────────────────────┘
               │ Policy YAML + Proxy
┌──────────────▼──────────────────────┐
│   Veea Lobster Trap (Port 8080)     │
│   Deep Prompt Inspection Proxy      │
│   YAML Policy Enforcement           │
└─────────────────────────────────────┘

---

Tech Stack

Layer	Technology
Frontend	Next.js 15 (App Router), TypeScript, Tailwind CSS
Backend	Python + FastAPI, uv
AI	Google Gemini 2.0 Flash
Security	Veea Lobster Trap (MIT)
Database	SQLite via SQLModel
Deploy	Vercel (frontend) + Railway (backend)

---

Local Setup

Prerequisites

• Python 3.11+ with uv
• Node.js 18+
• Google Gemini API key (free at aistudio.google.com)

Backend

cd backend
cp .env.example .env
# Add your GEMINI_API_KEY to .env
uv run uvicorn app.main:app --reload --port 8000

Frontend

cd frontend
npm install
npm run dev

Open http://localhost:3000

Lobster Trap (Optional — Mock mode enabled by default)

cd lobstertrap
docker compose up

Set MOCK_LOBSTERTRAP=false in backend/.env to use real enforcement.

---

Environment Variables

GEMINI_API_KEY=your_key_here
LOBSTERTRAP_URL=http://localhost:8080
MOCK_LOBSTERTRAP=true          # false = real Lobster Trap enforcement
DATABASE_URL=sqlite:///./policyforge.db
CORS_ORIGINS=http://localhost:3000

---

API Endpoints

Method	Path	Description
POST	/api/policies/generate	NL → YAML via Gemini
POST	/api/policies/save	Save + activate policy
GET	/api/policies	List all policies
DELETE	/api/policies/{id}	Deactivate policy
GET	/api/audit/logs	Recent audit events
GET	/api/audit/metrics	Dashboard metrics
GET	/api/audit/report?standard=HIPAA	Compliance report
GET	/api/demo/attacks	Available attack types
POST	/api/demo/attack	Fire a test attack

---

Hackathon Info

• Event: TechEx Intelligent Enterprise Solutions Hackathon
• Track: Track 1 — Agent Security & AI Governance
• Powered by: Veea Lobster Trap + Google Gemini
• Team: Muhammad Yaseen
• Deadline: May 19, 2026 "# policyforge"