Skip to main content

VulFinder is an autonomous multi-agent security engine that hunts website vulnerabilities, verifies exploits, and generates code-level fixes in real time.
Built with Next.js, FastAPI, and WebSockets, it replaces traditional static vulnerability scanners with an active AI testing taskforce:
• PathFinder: Controls a headless browser to map routes and intercept hidden API hooks.
• ThreatVector: Contextually crafts precise adversarial payloads targeting OWASP Top 10 flaws.
• VulnDoctor: Validates true threats using the CVSS framework to eliminate AI hallucinations.
• PatchMechanic: Generates ready-to-deploy remediation code patches for developers.

Built with

reactjavascripttailwindreact-native

Prize Categories Entered

  • DigitalOcean

    Best Open-Source AI Project

    Submitted