AAP / OAuth agent profiles
Profiles for OAuth-style delegation and accountable authorization in agent systems.
Doctrine reference for agent authority
Verifiable delegated authority for autonomous agents.
Agent Authority names the control layer between identity and action: who granted an agent power, what scope was delegated, how that scope is proven, and how it can be revoked.
Category Map
The authority layer is not only identity, and it is not only authorization. It is the accountable chain that makes an agent's action legible to systems, auditors, counterparties, and users.
IdentityWhich agent, principal, service, or organization is acting.
AuthorizationWhat the agent is permitted to do, for whom, and under what conditions.
DelegationHow authority moves from principal to agent without losing accountability.
ProvenanceWhich claims, credentials, policies, and signatures support the action.
RevocationHow authority is narrowed, expired, suspended, or withdrawn.
Standards Watch
The naming field is still forming. Current work around agent authorization profiles, agent identity protocols, accountable delegation, and machine-readable agent cards points toward a shared problem: authority must become inspectable infrastructure.
AID / AIP identity drafts
Protocol work around agent identity, verifiable delegation, and cross-domain claims.
MCP / A2A / agent cards
Discovery and tool surfaces need authority metadata, not only endpoint metadata.
x401 / payment authority
Agent-native payment flows require delegated intent, spending scope, and revocation semantics.
Reference Position
Agent Authority is maintained as a neutral doctrine surface for governance, control-plane design, registries, trust layers, and standards alignment. It is not a marketplace lander and does not present itself as inventory.
Collaboration, standards, and reference-surface inquiries can be routed through the quiet inquiry page.