Ask three questions in order
First, check evidence. Does each factual recommendation trace to the supplied notes, or does the draft turn a preference into a general claim?
Second, check fit. Does the artifact serve the named reader, format, and decision? Accurate material can still be wrong for the job.
Third, check scope. Did the draft introduce a channel, audience, promise, or deliverable that the task never requested?
The input is a draft artifact plus its task and sources. The output is a short finding that cites the affected section, explains the issue, and requests a bounded correction. Stop after every finding has a clear disposition in the revised artifact.
Apply the method to fabricated source notes
The evidence check compares “Saturday sessions are available” with the workshop note and finds no supporting schedule. The fit check confirms that the brief still serves independent consultants. The scope check keeps the two-hour offer and confirmed-detail constraint while excluding price.
Correction removes the Saturday sentence from the final brief. The checkpoint preserves the source, first brief, finding, correction, and unknown schedule and price.
This synthetic sequence demonstrates the three review questions, not a market outcome.
Use the method where the source set is visible
This review fits briefs, comparisons, operating notes, and other artifacts with a declared source set. It is a poor fit when the task has no boundary or the reviewer cannot inspect the evidence.