Protocol
Turn a production finding into a reusable eval artifact
A failure is expensive twice if the system can forget it.
When it fits
- A production incident teaches the team something, but the lesson is likely to disappear after the immediate fix.
When to avoid it
- Do not convert every noisy observation into permanent rules. Validate the finding and avoid leaking production secrets or personal data into repositories or eval sets.
Why it matters
Preserve a validated production finding in the form that best prevents recurrence: an issue with evidence, a regression example, an evaluator case or a dataset item. Keep the link to the original trace and the execution boundary where the failure mattered. Then rerun the relevant eval after the fix.
Steps
- Validate the finding against the original production evidence.
- Minimize sensitive data before storing a durable artifact.
- Choose the artifact that can catch the failure again.
- Link it to the affected component and original evidence.
- Rerun the relevant evaluation after the repair.
An example
A coding agent repeatedly edits generated files instead of their source templates. Preserve one verified incident as a regression case and repository instruction test before merging the fix.
Check your result
The next version is tested against an artifact derived from the real failure, and the source evidence remains traceable.
Keep this limit in mind
- Do not convert every noisy observation into permanent rules. Validate the finding and avoid leaking production secrets or personal data into repositories or eval sets.
Connected ideas
Useful withTurn a verified AI failure into a regression test
Evidence and sources
A production investigation can preserve evidence by turning a validated finding into a reviewable issue, evaluator or dataset example before the repair is automated.
Runtime evidence can contain noise, secrets or personal data; promotion into a durable artifact requires review, minimization and a clear link to the original failure.
From Signal to PR: Anatomy of a Self-Improving Agent · 3:31-18:04, gather evidence, bring it into the repository, create an issue/evaluator/dataset example and preserve the execution boundary