What should a lab agree before accepting a routing sample?

Use this checklist to establish a review process before assessing a generated sample, so acceptance reflects the intended research use and known limits remain visible when the scope expands.

- Name the buyer roles responsible for assessing operational plausibility, technical usability and research relevance, with a clear owner for reconciling conflicting review conclusions. - Define how reviewers will record disputed assignments, including the request reference, supporting policy and whether the issue concerns content, evidence or evaluator behaviour. - Specify which defects require sample correction and which identify an agreed scope limit, so unresolved problems are not silently converted into accepted behaviour. - Agree how corrected records and their dependent targets will be identified, allowing reviewers to recheck affected material without repeating an unrelated full review. - Require a coverage statement describing included decision patterns and exclusions, giving downstream researchers enough context to avoid overstating what the sample represents. - Decide what evidence would support expansion into another domain or routing boundary, rather than treating acceptance as approval for every adjacent workflow. - Preserve the review outcome with the accepted sample version, including limitations that must accompany internal handover to the teams using it.