Overview

TAP is a simple but powerful idea. It defines how AI credentials should be created, verified, revoked, and enforced. Each AI agent carries a verifiable credential, a kind of digital passport, that can be checked instantly by anyone using cryptographic proofs.

TAP replaces static, centralized checks with programmable trust.

It verifies that:

• The agent’s identity is real and issued by an approved source

• The credentials haven’t expired or been revoked

• The agent’s actions follow agreed-upon rules

Behind the scenes, TAP uses signatures, proofs, and attestations to confirm every step.

This turns trust from something you assume into something you can verify.

By connecting verifiable credentials, attested compute, and transparent rules, TAP creates a full trust cycle: agents prove who they are, complete a verified task, and record their results for audit or compliance.