Governance behavior for AI agents governed by DashClaw. Teaches the governance protocol: when to call guard (risk thresholds), how to interpret decisions (allow/warn/block/require_approval), when to record actions, how to wait for approvals, and session…
原文语言:英语