couldbeme/holdline

Holdline — a neutral benchmark for AI-agent write-guards. Catch rate, false-block rate, class-balanced kappa, and an injection-attack class. Scores any guard.

2 best points · 0 comments · 1 Hacker News post · 1 GitHub stars · View on GitHub

Hacker News discussions

• Show HN: A benchmark for AI agent guardrails that caught my own plugin (2 points, 0 comments, by couldbeme_, 2026-08-17)