Problem

The A2A protocol recommends TLS certificate validation and allows optional JWS signatures on AgentCards for agent identity. These mechanisms verify the domain hosting an agent but do not verify the agent itself. This creates security gaps in multi-agent ecosystems:

1. No verified agent identity - AgentProvider fields are self-asserted with no verification
2. No trust evaluation framework for agents
3. No trust propagation or authorization boundaries in delegation chains
4. No AgentCard revocation mechanism
5. No application-layer message integrity

Proposal

An extension-based identity framework using existing AgentExtension and metadata fields, with three verification levels (self-asserted, domain-verified, organization-verified), standardized trust signals, mandatory AgentCard signing for production, revocation endpoints, delegation chain security, and message signing.

Full proposal: docs/proposals/agent-identity-trust-framework.md
Design rationale: docs/proposals/agent-identity-trust-framework-design-rationale.md