Show HN: Open-Kritt – Open-source infrastructure for AI-based security research (github.com)

Open-Kritt has launched an innovative open-source platform designed to enhance AI-driven security research by enabling users to orchestrate AI agents to uncover real vulnerabilities in code efficiently. Unlike traditional methods that struggle with large-scale analyses, Open-Kritt adopts a focused approach by breaking down research into manageable tasks, allowing multiple AI agents to operate in parallel. The resulting findings are de-duplicated, ranked, and can be validated through reusable security research playbooks, making it a significant tool for security researchers and developers who seek a customizable workflow and model management. This platform is particularly noteworthy for the AI/ML community as it allows users to leverage AI models like Codex or Claude Code for comprehensive analysis of both remote and local repositories. It offers features such as custom severity ranking, automatic de-duplication, and the ability to build proofs of concept. With its roots in successful bug-bounty efforts totaling over $1.5 million, Open-Kritt stands to bolster security practices by providing a self-hosted, flexible alternative to existing tools. Developers can easily set it up using Git and Docker, ensuring accessibility while maintaining a focus on securing the research environment.