What You Need to Know
Reviewing an output means checking what Claude wrote before anyone relies on it. A review asks two questions. Accuracy asks whether the specifics are right: numbers, names, dates, titles, quoted figures, and references to documents or rules. Completeness asks whether anything is missing. A review checks both against the source. The source is the material that Claude used to build the output.
This lesson uses one made-up example throughout: a property management company. An office manager attaches a cleaning supplier's new contract and asks Claude for a summary. Her director will use the summary to decide whether to sign. The summary reads well, but it could contain a wrong monthly price or omit a clause.
The Reviewer's Stance
Treat every Claude output as a good first draft that can contain errors stated with confidence. The Help Center says why: Claude "can occasionally produce responses that are incorrect or misleading", and users "should not rely on Claude as a singular source of truth". Fluency and confidence are properties of the writing, not evidence about the facts. Claude writes fluently whether it is right or wrong, so an output that reads beautifully has passed no check. The contract summary would read just as well with a wrong price.
Ownership decides who does the checking. The professional who sends Claude's work owns it. That professional is accountable for it to their team, their manager or their client. Review is the step that changes a draft into work that you can accept responsibility for. Review is not an optional extra for cautious people.
Accuracy: Are the Specifics Right?
Check the specifics first, in order of consequence. Start with the details that would be most costly or embarrassing if they were wrong. These are usually the figure that a decision depends on, the name at the top of the letter, and the date of the deadline. A draft with twenty accurate sentences and one wrong subtotal is a wrong draft if the meeting decides on that subtotal. In the contract summary, the monthly price comes first, because the director decides on it. Some figures, such as a yearly cleaning cost, are in no clause of the contract, because Claude calculated them. Objective 2.3 explains how to check a calculated figure.
Completeness: Is Anything Missing?
Completeness is harder to check, because the evidence is absent rather than wrong. A fluent summary can omit a section, a theme or a caveat, and nothing in the prose shows the gap. So check the output against the source itself, never against your memory of the source. Follow the source's own structure: its sections, funds, themes or line items. Confirm that each element is either in the output or excluded on purpose. If you read the output alone and accept it, you send it with the gaps still in it. The office manager ticks each clause in the contract against the summary.
Check accuracy first every time. A wrong specific is the most likely failure, and it is the quickest to catch. Check completeness second, because that check needs the source open anyway.
Verify Against the Input
When Claude builds an output from attached material, the first check is the attachment. The attachment is already in the conversation, and it takes minutes to compare the two. Anthropic's prompting guidance gives the same advice for long documents: ask Claude to quote the relevant parts of the source before it does the task. In the chat, ask Claude where in the attachment each figure or claim came from. Then spot-check the claims that matter against those locations. An output whose claims trace to the source is checkable in minutes. Do not assume that an output is correct when you cannot trace its claims. The office manager asks where the price came from, then reads that clause.
Asking Claude whether it is sure does not confirm anything, because no new information enters the check. Claude can miss its own error again and still answer yes. A real check compares the output with its source. Objective 2.3 explains what a self-check can find, and why it never replaces that comparison.
In the exam
This domain has a larger share of the exam than any other domain, and every objective in it depends on the reviewer's stance. The certification's central claim is that its holder can be trusted with what Claude produces. The holder earns that trust at the review stage, not at the prompting stage. When the stance is right, the other skills in the domain follow from it: spotting failure types, validating claims, knowing when to escalate. The exam treats self-reported confidence as a wrong answer, never as a check.
The sample question in the exam guide
The guide's Domain 2 sample question tests this objective. Claude summarises a regulation and cites a specific subsection. The right answer is to verify that citation against the official text before the compliance team sees it. The named wrong answers are trusting Claude's expressed confidence and rewording the summary to sound more formal. Diligence matters more than confidence, and polish is never a substitute for correctness.
Exam Traps
Sending the output because it reads well or because Claude expressed high confidence
Fluency and confidence are properties of the writing, not evidence about the facts, so neither has checked anything. Compare the output against the source material before anyone relies on it.
Asking Claude to confirm its own work and treating a confident yes as review
No new information enters the check, so Claude can confirm its own error with the same confidence. Compare the output with its source, or ask Claude where in the source each claim came from and spot-check those locations.
Verifying one named figure when part of the source may have been dropped
Accuracy checking alone misses the second review question. When the source had more in it than the output shows, review each part of the source's own structure for completeness, not just the details that you noticed.
Practice Scenario
A finance analyst at a bank has Claude draft the quarterly spending summary for a management meeting from the bank's budget spreadsheet. The draft reads perfectly, but she must be sure it is right and leaves nothing out, without simply trusting what Claude says. What should the analyst do before circulating it?
Build Exercise
Run the Two-Pass Review
What you'll learn
- How to check every specific in an output against the source material
- How to order accuracy checking by consequence
- How to walk a source structure to catch omissions
- How to make an output traceable by asking where each claim came from
- Take a document that Claude drafted for you from real source material, such as a summary, a report or a data write-up. Or generate one now from a spreadsheet or document you actually use
Why: The habit of checking accuracy and then completeness shows its value only on real work, where you know what the source contains and what a miss would cost.
You should see: A Claude draft and its source material side by side in one conversation.
- Accuracy check: list every specific in the output. Specifics are figures, names, dates and references. Check each against the source, starting with the ones a decision or reader depends on
Why: Ordering by consequence puts the limited checking time where an error would cost most.
You should see: A marked-up list of specifics: confirmed, wrong, or unverifiable.
- Completeness pass: review each part of the source’s structure. The structure is its sections, funds, themes or line items. Confirm that each element is represented in the output or deliberately excluded
Why: Omissions are not visible in the prose; only the source’s structure shows what should be there.
You should see: A tick against every element of the source, or a note explaining its absence.
- Ask Claude to indicate where in the source each remaining claim came from, and spot-check the important claims
Why: A traceable output can be verified in minutes; an untraceable one is treated as unverified.
You should see: Each claim tied to a location in the source, with the important ones confirmed.
- Record what the two checks caught and roughly how long they took
Why: You keep the habit when you can see its cost and its benefit on your own work.
You should see: A short note: errors caught, omissions caught, minutes spent.