Subject: Meeting Summary – AI Agent Protocol CG (2026-06-17)

Dear all,

  Apologies for the slight delay in sharing the meeting summary.

  The minutes from the AI Agent Protocol CG meeting held on June 17, 2026,
have now been compiled. Please review them at the link below and feel free
to provide any
  feedback:

  Meeting minutes: https://github.com/w3c-cg/ai-agent-protocol/blob/main/
meetings/2026-06-17.md

  This meeting focused on follow-up discussion from the previous
presentation on updates to the CG use-case and protocol documents,
including DID-based agent identity,
  discovery, and the planned OAuth/OIDC integration path.

  Sean also presented a proposed model for how agents can understand
network state, authentication, authorization, permission challenges, and
valid next actions. The
  discussion covered agent-to-agent permission flows, agent-to-resource
access, and the need for structured interaction-state, action, and
authorization-recommendation
  formats.

  The group also reviewed the plan for resolving GitHub issues before
finalizing the protocol and use-case documents as first-year CG
deliverables.

  Protocol draft: https://w3c-cg.github.io/ai-agent-protocol/protocol.html
  GitHub discussion: https://github.com/w3c-cg/ai-agent-protocol/

  The meeting closed with discussion of the W3C/GS1 Workshop on E-commerce
for Humans and AI Agents, including remote participation, position
statements, and Expressions
  of Interest.

  Workshop: https://www.w3.org/2026/ecommerce-agents/

  We welcome any comments on the minutes or on the topics discussed.

  Best regards,

  Song and Gaowei