Trusted Agent Protocol (TAP)

A dated reading of what is claimed, reported, and independently verified in the current evidence.

Current reading

The readiness gap, in one scan

AI-assisted assembly · derived results

Evidence strength Critical

Decision

Human editorial judgment · 2026-06-28

Adopt with guardrails

Why: TAP is built on a real IETF standard (RFC 9421), is openly specified under MIT, has 100+ enrolled partners and a multi-PSP Oct 2025 launch, and is already part of our protocol stack — making it the most adoptable of the three. Guardrails needed because the reference impl is demo-grade (6 commits, no releases) and trust is rooted in a Visa-operated key directory.
Next: Since TAP is already a named layer in our protocol architecture, validate our TAP identity layer against the published RFC 9421 spec (Ed25519/PS256 signatures, 8-minute validity window, nonce replay protection, agent-browser-auth vs agent-payer-auth tags) and align our header/JWT schema to Visa's; keep our implementation directory-agnostic so we are not hard-bound to Visa's JWKS.

Constraints

No named blocker is present in the current public projection.

Evidence summary

AI-assisted assembly

Counts and dates only. Raw signals, private excerpts, trust records, and internal corpus material are not published here.

Publication record

Initial public reading

This is the initial public reading. No earlier readiness change is recorded.