HOL Guard
Free, local guard that reviews every AI coding agent action before it runs
HOL Guard is a local-first runtime security layer for AI coding agents. It reviews risky actions before they run — destructive commands, secret and password access, shady installs, and data exfiltration — and lets you allow, ask, or block each one under your own policy, keeping a local record of every decision. It works with Codex, Claude Code, Cursor, OpenCode, Antigravity CLI, and other agents through their hooks, runs entirely on your machine with no cloud account, and keeps working offline. The core Guard Local runtime is open source under Apache-2.0; Guard Cloud is an optional paid add-on for synced history, shared policy, and team dashboards (Solo from $4.99 per month after a 3-day trial).
سجّل الدخول لاقتراح بديل لهذا المنتج.
تسجيل الدخول