HAL – Harmful Action Limiter: Lean command guard for AI coding agents.

(github.com) by otherland26 | Mar 13, 2026 | 0 comments on HN