action-enforcement

Tag

Cards List
#action-enforcement

OpenAgentFlow: Enabling System-Wide Safety Boundaries for Heterogeneous AI Agent Fleets

arXiv cs.AI · 2026-09-02 Cached

OpenAgentFlow presents a new architecture for system-wide safety enforcement in heterogeneous AI agent fleets, using a unified action stream and pre-execution policy enforcement to ensure safe operation.

0 favorites 0 likes
← Back to home

Submit Feedback