works on
action verification 1agentic AI 1policy enforcement 1runtime governance 1semantic pattern detection 1

From the 1 of 8 linked papers with an AI index.

activity
20242026
collaborators
Showing cs.AIShow all

1 paper · 1 filter

cs.AI2026

CAVA: Canonical Action Verification and Attestation for Runtime Governance of Agentic AI Systems

Zexun Wang

The paper introduces CAVA, a runtime layer that converts diverse agent actions into a unified, verifiable canonical form, enabling clear approval evidence and reproducible verifica…