jimmyracheta/AI-Runtime-Guard
Developer ProductivityA policy‑enforced guard that intercepts every file system and shell action of AI agents before execution, blocking destructive commands or requiring human approval.
Features
- Blocks dangerous commands (e.g., `rm -rf`) before they run
- Requires explicit human approval for risky operations via a local GUI/API
- Enforces workspace‑level path boundaries and per‑agent policy overlays
- Logs all actions to an activity log and indexes them in a dashboard database
Recent releases
View all 13 releases →
Review required
v2.3.0
Breaking risk
Auth
RBAC
RCE / SSRF
+1 more
Security hardening + UI fixes
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Install & Platforms
Install via
pipx
shell-script
Platforms
macos
linux
