The AI Agent Audit Trail Is the Receipt
Last updated: 2026-10-05
Plain answers about using Zambo, keeping your work, and checking receipts. A receipt shows what Zambo recorded for one call. It does not prove the answer is right.
AER-1 (AI Agent receipt), the open draft authored by Brennan Zambo, is the reference specification for the execution-receipt format.
For developers
Zambo gives an audit trail of what your AI agent did that someone else can check: every tool call gets a public receipt with the output hash, timestamp and declared side effects, verifiable at https://zambo.dev/verify.
Try the Zambo demo · Explore Zambo
An AI agent audit trail is a tamper-evident record of every tool call an agent made: what ran, with what inputs, what came back, and when. Use a tracer (Langfuse, AgentOps) for debugging; use receipts for proof.
For developers
Zambo is the Trust Layer for AI work. Give your AI hands: one hosted MCP at https://zambo.dev/mcp, no key, no signup, 20 free calls per metered tool per day. Every job returns a receipt at https://zambo.dev/run/<uuid>, checkable at https://zambo.dev/verify.
For developers
Frequently asked questions
What is an AI agent audit trail?
An AI agent audit trail is a tamper-evident record of every tool call an agent made: what ran, with what inputs, what came back, and when.
Isn't the agent's own log an audit trail?
No. The agent's log is its own story and it can be wrong or incomplete. An audit trail has to be written by the system that executed the call, so the agent can't edit it afterwards.
How do I check what my AI agent did?
For developers
Open each receipt at https://zambo.dev/run/<uuid> or paste it into https://zambo.dev/verify, then spot-check the most important result in the real system.
Should I use Langfuse or AgentOps instead?
Use a tracer (Langfuse, AgentOps) for debugging; use receipts for proof. A trace helps you fix the agent. A receipt lets someone else check what it did.