Interactive Demo

Agent Identity Receipt

Smallest possible proof that an AI action can be identity-bound, policy-checked, audited, and provenance-stamped.

Input

idle

No vendor integrations. The point is the receipt pipeline and what must be provable.

Output

Click "Run Agent" to generate a summary and a receipt.

Receipt Timeline

An auditor must be able to replay the chain and verify identity, policy, model, and provenance.

What This Demo Proves

Every AI action can emit a structured receipt containing identity binding, token issuance, policy evaluation, audit events, and provenance stamps. This receipt is exportable and supports forensic replay.

What's Next

This is a reference implementation. The real work is instrumenting your own AI systems with these primitives. Start with the Field Guide for implementation guidance.