Dangerous Command Block

hookAI & models

Blocks your agent from running destructive commands like file deletion or database drops.

One `rm -rf`, force-push, or `DROP TABLE` from an over-eager agent is unrecoverable. The harness should refuse destructive commands rather than trust the model to be careful.

Delivery for this kind is on the roadmap — not serving yet. You can still add it. It stays paused until ahel can serve it.

Serve it through your gateway

One link, every agent. Your own credentials, stored once.

Signals

GitHub stars
4k
Forks
374
Last commit
Aug 2026