construct-auto-classifier

A safety gate that sits in front of AI coding assistants' shell-execution tools (Antigravity, OpenCode, and others), first applying fast structural rules and then having Jev or any chat LLM judge what a borderline command actually does — reported as certified with zero dangerous commands allowed through in testing.

unverified — sourced from awesomejev.com; description independently written by AionEdge from the repo’s own GitHub metadata and README, not GitHub-API relevance-checked