This repository contains research code and data for evaluating AI-security scanners. The corpus is analyzed statically; repositories in the study are not executed by the benchmark.

If you discover a security vulnerability in the research tooling itself, please report it privately through GitHub's Security tab using Report a vulnerability, rather than opening a public issue.

Include:

• the affected file, component, or workflow
• a minimal reproduction or proof of concept
• the security impact
• any relevant environment or version information

Please do not disclose credentials, API tokens, or other sensitive data in a public issue.

False positives, scanner-quality findings, methodological concerns, and reproducibility issues are welcome as normal GitHub issues unless they expose an exploitable vulnerability in the repository itself.

Security fixes target the default branch. Historical commits and research snapshots are not maintained as supported releases.