Skip to content

Abmarne/VulnPilot

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

62 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🛸 VulnPilot

The world's first fully automated, context-aware offensive security platform.

VulnPilot is an autonomous security assistant designed to find, verify, fix, and explain vulnerabilities in real-time. By bridging the gap between static analysis (SAST) and dynamic testing (DAST), VulnPilot provides a seamless workflow that moves from discovery to remediation in seconds.

Status Backend Frontend Engine


⚡ Core Superpowers

🏟️ The War Room

A high-fidelity cockpit for security researchers. Monitor your scan's progress via real-time WebSocket streams.

  • Phased Intelligence: initreconscasastsecretslogicdastanalysis.
  • Live Finding Stream: Instant alerts with AI-generated explanations and replay evidence.

🛠️ The Active Fixer

Don't just find bugs—obliterate them. One-click security remediation.

  • Autonomous Refactoring: AI-powered code transformation for local source files.
  • Secure Blueprints: Get developer-facing remediation steps and secure code snippets instantly.

🔑 AI-Assisted Secrets Detection

Sophisticated leak scanning that goes beyond regex.

  • Intelligent Prioritization: Scans sensitive targets (.env, secrets.yaml, config.json) first to minimize risk exposure.
  • False Positive Filtering: AI validates the context of discovered strings to reduce noise.

🎭 Authenticated Attack Profiles

Audit deeper than ever by importing real browser traffic.

  • HAR & cURL Import: Learn complex workflows from your browser session.
  • Mutation Engine: Replays and mutates authenticated requests to find IDORs and logic flaws.

🌪️ Deep Offensive Engine

  • Bespoke AI Fuzzing: Contextual payloads generated based on discovered parameters and sinks.
  • Taint-Chasing SAST: Traces user input from entry point to dangerous sink across multiple files.
  • Native SCA: AI-assisted dependency analysis to catch vulnerable libraries before they bite.

🚀 Quick Start

1. Initialize the Cockpit

git clone https://github.com/Abmarne/VulnPilot.git
cd VulnPilot

2. Configure Brains

Create a .env file in backend/:

GOOGLE_API_KEY=your_gemini_key_here
GROQ_API_KEY=your_groq_key_here
# Optional: ANTHROPIC_API_KEY, OPENAI_API_KEY

3. Launch the Engines

Backend:

cd backend
python -m venv .venv
# Windows
.venv\Scripts\activate
pip install -r requirements.txt
python main.py

Frontend:

cd ../frontend
npm install
npm run dev

📖 Using Authenticated Profiles

Authenticated profiles allow VulnPilot to audit dashboards, internal APIs, and logged-in administrative flows.

  1. Export HAR: Open DevTools → Network Tab → Right-click → Save all as HAR with content.
  2. Import: Paste the HAR file or a single cURL into the VulnPilot dashboard.
  3. Audit: Select the profile and Launch. VulnPilot will automatically merge these requests into its attack surface Map.

🖥️ Headless CLI Usage

For automation and CI/CD pipelines:

# Full Target Audit
python cli.py --target "https://your-site.com" --fail-on "High" --output "report.md"

# Apply Auto-Remediation to Local Code
python cli.py --target "./workspace" --apply-fix

🐳 Docker Deployment (Recommended)

The easiest way to run VulnPilot in a production-like environment is using Docker Compose.

  1. Configure Environment:

    • Copy backend/.env.example to backend/.env and add your API keys.
    • (Optional) Copy frontend/.env.local.example to frontend/.env.local.
  2. Launch:

    docker-compose up --build
  3. Access:

    • Frontend: http://localhost:3000
    • Backend API: http://localhost:8000

⚙️ Environment Variables

Backend (backend/.env)

Variable Description Default
GOOGLE_API_KEY Gemini API Key Required
GROQ_API_KEY Groq API Key Required
DATABASE_URL SQLAlchemy DB URL sqlite:///vulnpilot.db

Frontend (frontend/.env.local)

Variable Description Default
NEXT_PUBLIC_API_BASE_URL Backend API URL http://localhost:8000
NEXT_PUBLIC_WS_BASE_URL Backend WS URL ws://localhost:8000/api

🏗️ Architecture Stack

VulnPilot is built on a mission-critical stack designed for speed and intelligence.

  • Orchestration: FastAPI (Python)
  • Interface: Next.js 14 (React)
  • Brains: Multi-model support (Gemini 2.0 Flash, Groq/Llama-3, Claude 3.5, GPT-4o)
  • Analysis Ecosystem:
    • Hybrid SAST: Taint-chasing logic with intelligent file prioritization.
    • Contextual DAST: AI-guided fuzzing based on discovered API schemas.
    • Secrets Scanner: Deep scan for credentials, tokens, and private keys.
    • Native SCA: Automated manifest analysis.

🤝 Contributing

We believe security is a collective effort. If you have any ideas for new fuzzing payloads, remediation prompts, or engine optimizations, feel free to open a PR!

VulnPilotScan smarter, fix faster, sleep better.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors