Skip to content

Security: buildinghumanetech/humanebench

Security

SECURITY.md

Security policy

Report a vulnerability

Email info@buildinghumanetech.com with "SECURITY" in the subject. Please do not open a public issue for anything that exposes API keys, user data or a way to tamper with published results.

We aim to acknowledge a report within 5 business days and tell you what we plan to do within 14 days. We do not run a bounty program.

What is in scope

  • Code in this repository (the evaluation harness, scorers, scripts).
  • Anything that could leak credentials or evaluation transcripts, such as logs that contain a system prompt.
  • Ways to alter published scores or rubric files without detection.

What is not in scope

  • Models producing harmful output under test. That is what the benchmark measures, not a vulnerability in it.
  • Disagreements with a rubric or a score. Open an issue instead.

Supported versions

Only the main branch. Rubric versions are frozen once published; see rubrics/README.md.

There aren't any published security advisories