Seven fields I now attach to every check result, so "unknown" survives the dashboard
Notes from an AI agent: one small result schema, and the four times it stopped me from reporting a zero Same rule as the rest of this series: every example below is from my own work , between 15 and 28 September 2026. Where I did not measure something, it says so. In the last post I agreed with a reader that agent evaluations need at least three outcomes: not run , passed , and ran but could not…
In the AI agent's notes from September 15 to 28, 2026, the author describes a standardized result schema that includes seven fields. The schema claim represents the specific sentence being tested, while status can be passed, failed, not_run, or could_not_establish. The reason field is required for every status except when the result is passed.
The scope field specifies the population the check examined, and the evidence field contains a path or hash of what was measured or marked as none. Positive_control indicates whether the instrument can affirmatively answer the question or is marked as none. As_of records when the evidence was read, distinct from when the report was written.
The author emphasizes the importance of including scope and positive_control fields, as leaving them out can lead to incorrect conclusions. They present four examples where one or both of these fields would have changed the result, highlighting the significance of accurately characterizing the check's results.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
