Anthropic's open-source framework for AI-powered vulnerability discovery
Anthropic just open-sourced a framework that turns Claude into a vulnerability hunter. Security teams can deploy it today.

Why it matters
Anthropic is moving beyond chat into developer tooling by releasing a production-ready framework that leverages Claude's reasoning capabilities for code security. This signals a shift toward AI-as-infrastructure for enterprise security workflows.
The key facts
6 to knowOpen-source framework for AI-powered vulnerability discovery
Built on Anthropic's Claude model
GitHub repository: defending-code-reference-harness
Published June 4, 2026
Community engagement: 38 HN points, 13 comments
Targets developer/security team workflows
Go to the source
Hacker Newsgithub.com
Publisher excerpt: Article URL: Comments URL: Points: 38 # Comments: 13