Fail-closed execution boundary for consequential AI-agent actions.
autonomous-agents ai-agents policy-enforcement tool-calling fail-closed agent-safety pre-execution interai
-
Updated
Sep 16, 2026 - JavaScript