Why a short handoff brief before a coding agent starts prevents wasted review

Agent runs often go wrong not because of the code but because the task's intent never fully reached the prompt. Writing a brief that states the goal, scope, off-limits areas, checks, and completion criteria turns review into a simple comparison against what was asked. Ambiguities should be listed in advance under an escalate section so the agent surfaces them as questions rather than guessing silently. Durable project rules belong in the instruction file the tool already reads, leaving the brief focused on the single task. The author illustrates this with a deprecated client-method migration, where a precise brief answers the questions a vague one leaves open. Reviewers should check the diff against the brief's scope and rerun the checks themselves rather than trusting a summary.