The open protocol for bounded AI-agent authority: humans authorize, agents execute, and every consequential action requires a signed receipt before it runs.https://humanagencyprotocol.org/enhttps://humanagencyprotocol.org/governancehttps://humanagencyprotocol.org/governanceHow HAP is governed: invariant constraints instead of institutions. No central authority, no registry, no approval process — conformance is enforced locally.Mon, 01 Jun 2026 00:00:00 GMThttps://humanagencyprotocol.org/reviewhttps://humanagencyprotocol.org/reviewNon-normative v0.5 material: optional extensions and future directions an implementation may adopt or skip without affecting conformance.Mon, 01 Jun 2026 00:00:00 GMThttps://humanagencyprotocol.org/protocolhttps://humanagencyprotocol.org/protocolThe normative HAP specification: profiles, bounds, gates, attestations, and the signed receipt an agent must obtain before a consequential action runs.Mon, 01 Jun 2026 00:00:00 GMT