What is an agent execution receipt standard?
The short answer An agent execution receipt standard is a published specification that defines, for a record of one AI agent tool call, four things exactly: what the record contains, how the commitment over its contents is computed, how an independent checker verifies it and what outcomes that checker is allowed to report, and a conformance corpus that proves two implementations built separately…
An agent execution receipt standard is a specification that outlines the essential details of a single AI agent tool call. It specifies what the record includes, how its commitment is computed, how an independent checker verifies it, and what outcomes the checker can report. This standard is exemplified by AER-1, published as IETF Internet-Draft draft-zambo-aer1-26.
The receipt record contains four main components: the execution details, the commitment over the output, the verifier's outcomes, and a conformance corpus. These components ensure the record's integrity, identify the executed action, and provide a method to confirm the results. The standard also separates verified data from sources, enabling independent verification without trusting the person presenting the record.
The specification separates the saved result from claims about the outside world, preventing silent upgrades of producer assertions into facts. Verification involves recomputing the SHA-256 hash over the canonical bytes and comparing it to the output_hash. AER-1's conformance corpus, kit 1.5.7 with 143 test vectors, ensures independent implementations produce matching results. A live receipt example demonstrates how each component functions in practice.
Rewritten by urgent.news; may contain errors — check the original.
Source: dev.to