Drop-in runtime security for AI agents. See every action in real time. Open rules. Free to start.
Know what your agents are doing, every action, in real time. No guessing.
The rules that protect your agents are open. Read them, fork them, run them yourself.
One command in your terminal. No account, no credit card, no sales call.
Basic is free forever. Upgrade to Advanced when you want attacks blocked automatically, not just flagged.