How to write a custom AI reviewer agent

A good reviewer agent brief has five parts: a one-sentence scope, an ordered list of specific checks, an instruction on what to do when information is missing, an instruction on what not to do, and a required output shape. Write the checks as things to determine rather than topics to consider, order them so the highest-value check comes first, and keep the whole brief under 6,000 characters. The most common mistake is writing a brief that is too broad.

This page is about writing review prompts generally. It applies whether you paste the result into PerfectPaper or use it elsewhere.

Step one: make the scope narrow

Write one sentence naming what the agent examines and what it ignores. An agent briefed to check statistics, figures and writing quality will produce shallow findings on all three, because attention is finite and the instruction gives it no priority.

Narrow scopes produce specific findings. “Check the manuscript for problems” produces the observations you could have made yourself.

Step two: write checks as determinations, not topics

The difference between a brief that works and one that does not is usually grammatical. Compare:

Consider whether the statistical analysis is appropriate.

Determine, for every reported comparison, whether the stated n counts animals or cells, and report every instance where the n exceeds the number of independent units.

The first invites an opinion. The second specifies an operation with a definite output, and can be checked. Number your determinations so they run in a defined order.

Step three: order the checks by value

Put the check most likely to produce a consequential finding first, and say so explicitly — an instruction like “treat this as the highest priority item” changes what gets attention. The last item in a long list gets the least.

Step four: say what to do when something is missing

Most review findings are absences, and absences are where agents guess. Include an instruction of this shape:

Where you cannot determine something from the manuscript, say exactly what is missing and what the author would need to add. Do not infer values that are not present.

Without it, an agent asked whether blinding was performed will sometimes report that it was, on the strength of the paper reading like a careful one.

Step five: write the negative instructions

The instructions telling an agent what not to do are the ones that keep a brief from producing confident nonsense. They are also the first thing people delete when shortening a brief, and they should be the last.

Examples from the prebuilt briefs on this site:

• Do not attempt to validate that a specific accession exists in any database. A false report of a missing accession is worse than not checking.
• Do not treat small sample size as a limitation in itself. A qualitative study evaluated by quantitative standards produces findings that are all wrong.
• Where the manuscript explicitly adopts a causal framework with stated assumptions, causal language is appropriate and should not be flagged. Without this, the agent flags every correct paper.
• Do not comment on grammar, formatting or citation style. Otherwise a specialist agent spends its attention on what four other agents already cover.

Step six: specify the output

State what each finding must contain. A useful default:

Report each issue with its exact location, what it prevents a reader from concluding, and a specific remedy.

The middle clause is the one worth insisting on. “Blinding is not reported” is a checklist item. “Blinding is not reported for caliper measurement, so the primary outcome may be biased by expectation” is a review comment.

A template

Name: [narrow, specific]

You are reviewing [scope]. Read [which sections] together.

Determine and report:

1. [Highest-value determination. Treat this as the priority item.]
2. [Determination, phrased as an operation with a definite output.]
3. [...]

Where you cannot determine something from the manuscript, say exactly what is
missing and what the author would need to add. Do not infer values that are
not present.

Do not [the specific failure mode this agent is prone to].

Report each issue with its exact location, what it prevents a reader from
concluding, and a specific remedy.

Settings that matter

Choose a work type matching the task — verification for checking work, statistics or causal inference for analysis, domain review for field-specific conventions. Choose expert skill level for anything requiring specialist knowledge. Add tools only where the task needs external information: a checking task that reads only the manuscript is usually more reliable without web search, because a tool available is a tool that gets used.

Browse the prebuilt agents by field for complete worked examples, all following this structure.

Review my manuscript

Frequently asked questions

How do I prompt an AI to peer review my paper?

Give it a narrow scope, an ordered list of specific determinations rather than topics, an instruction to say what is missing rather than infer it, explicit instructions on what not to do, and a required output shape naming location, consequence and remedy. Breadth is what makes a review prompt weak.

How long should a reviewer agent brief be?

Long enough to name every check specifically, and no longer. The prebuilt briefs on this site run between roughly 1,500 and 3,000 characters against a 6,000-character limit. Length spent on specific determinations helps; length spent on general encouragement does not.

Why do negative instructions matter so much?

Because the common failure of a review agent is not missing a problem but inventing one. Instructions like “do not infer values that are not present” and “do not treat small sample size as a limitation” prevent whole classes of false finding, and they cost a single line each.

Should I give my agent web search?

Only when the task genuinely requires information from outside the manuscript, such as comparing a compound’s effective concentration against reported human exposure. For checking tasks, tools add failure modes without adding accuracy.

Can I reuse the same agent across papers?

Yes, and it is worth building a small personal set. An agent written for the convention your last reviewer objected to is the most valuable one you will ever write, because that objection will come again.