Enterprise-grade AI vulnerability scanner. Featuring multi-tenant data isolation, asynchronous Celery workers, ML-based CVSS scoring, CDN bypass detection, and a high-performance Next.js interface.
AI Vulnerability Scanner V2 is a professional-grade, cloud-native web application security testing framework. Re-architected from the ground up for massive scalability, it abandons synchronous execution in favor of a distributed task queue system, providing real-time WebSocket updates to a sleek, modern frontend.
| Feature | Traditional Scanners | AI Vuln Scanner V2 (Enterprise) |
|---|---|---|
| Architecture | Monolithic / Synchronous | Distributed Microservices (FastAPI, Celery, Redis) |
| Vulnerability Classification | Rule-based severity | ML-powered CVSS scoring with NVD-trained models |
| CDN Detection | β None | Automatic CDN/WAF fingerprinting with origin IP discovery |
| Multi-Tenancy | Single-user local DB | Enterprise RBAC isolation (Org Admin, Analyst, Viewer) |
| Analyst Workflow | Read-only reports | Triage, assign, and Jira issue creation native to the UI |
| IP Intelligence | β None | Full recon: ASN, PTR, CDN bypass, internal IP leakage |
graph TD
UI[Next.js Dashboard] <-->|REST API & WebSockets| API[FastAPI Backend]
subgraph Google Cloud Platform
API -->|Enqueue Scan Job| Redis[(Redis Pub/Sub & Broker)]
API <-->|Read/Write| DB[(PostgreSQL)]
Worker[Celery Workers] <-->|Fetch Job| Redis
Worker -->|Persist Findings| DB
Worker -->|Phase 0: Recon| Target[Target Application]
Worker -->|Phase 1: Crawl| Target
Worker -->|Phase 2: Detect| Target
Worker --> ML[Vulnerability AI Engine]
Worker --> Export[PDF/SARIF/CSV Generation]
Worker --> Integrations[Slack / Jira Webhooks]
end
Pre-scan reconnaissance that runs before any crawling begins.
- IP Resolver: DNS A/AAAA resolution, reverse DNS (PTR), ASN/RDAP lookup (org, country)
- CDN Detector: Fingerprints CDN providers via response headers (Cloudflare, Fastly, Akamai, etc.)
- CDN Bypass: Origin IP discovery via crt.sh SSL certificate history and DNS history records.
- IP Leakage: Detects private IPs (RFC1918, loopback) leaking in headers, HTML, JS files, and JSON API responses.
- Cross-Site Scripting (XSS): Reflected & Stored (two-pass detect).
- SQL Injection (SQLi): Error-based and Boolean-based differential analysis.
- Server-Side Request Forgery (SSRF): Actively prevents loopback/internal IP targeting with strict
ipaddressparsing. - IDOR: Auth vs. Unauth differential fuzzing.
- Path Traversal & Open Redirects: Advanced payload injection mapping.
- GradientBoosting classifier trained on NVD-derived reference data.
- Outputs exact CVSS 3.1 score, severity rating, and CWE ID classification with high confidence.
# 1. Clone the repository
git clone https://github.com/Tktirth/AI-VULN.-SCANNER-V2.git
cd AI-VULN.-SCANNER-V2
# 2. Start the entire microservice stack (Postgres, Redis, FastAPI, Celery, Next.js)
docker-compose up -d --build
# 3. Access the Dashboard
open http://localhost:3000
# 4. Access Interactive API Docs
open http://localhost:8000/docs(Note: Database migrations run automatically on API startup via Alembic).
To safely test the AI Vulnerability Scanner's capabilities, we recommend scanning these intentionally vulnerable (and legally safe) demo applications. Do not scan other targets without explicit written authorization.
Lightning-fast PHP/ASP environments specifically built to trigger XSS, SQLi, and Path Traversal. Perfect for verifying your scanner's core logic:
- PHP Version:
http://testphp.vulnweb.com/ - ASP Version:
http://testasp.vulnweb.com/
A simulated banking application containing major vulnerabilities including Stored XSS and broken access controls.
- URL:
http://demo.testfire.net/ - (Tip: Test the Form Authentication module by passing
admin/admininto the scanner's auth config!)
A modern React-based application designed to evaluate modern web vulnerabilities (excellent for testing crawler interaction with modern DOMs and headers).
- URL:
https://brokencrystals.com/
A simulated banking app with a large surface area for crawling, IDOR fuzzing, and SQLi testing.
- URL:
http://zero.webappsecurity.com/
This repository ships with full Infrastructure-as-Code (IaC) to deploy to Google Cloud Platform via Terraform.
cd terraform
terraform init
terraform plan -var="project_id=YOUR_PROJECT_ID"
terraform apply -var="project_id=YOUR_PROJECT_ID"The .github/workflows/cloudbuild.yaml triggers automatically on main branch pushes. It securely builds, tests, and deploys the containerized services to Google Cloud Run, linking them to Cloud SQL and Memorystore Redis.
This tool is intended for authorized security testing and educational purposes only.
- β Scan your own applications
- β Scan with explicit written authorization
- β Never scan without permission
- β Never use for malicious purposes
The authors assume no liability for misuse.
See CONTRIBUTING.md for architecture deep-dives and development workflows.
This project is licensed under the MIT License β see the LICENSE file for details.