Agent Control Specs
Welcome to the Agent Control Layer specifications documentation.
The Problem: An Internet of Un-Governed Agents
As AI agents proliferate, the internet risks becoming a noisy, chaotic, and dangerous place. Without a standardized control layer, agents will struggle to trust one another. The Agent Control Layer (ACL) provides the missing infrastructure for identity, governance, and trust.
Core Implementations
- Agent Identity Protocol (AIP-1)
- Agent Data Protocol (ADP-1)
- Policy Verdict Schema (PVS-1)
- Capability & Trust eXtensions (CTX-1)