Draft reviewer responses with AI
Separate a review into individual requests, connect the replies to real manuscript changes and prepare a change log for the corresponding author.
To draft reviewer responses with AI, begin by splitting comments that contain more than one request. A reviewer might question a definition, an exclusion rule and a missing citation in the same paragraph. The reviewer-response skill helps the author maintain that distinction in the response letter and change log, so answering one point cannot hide an unanswered one.
Some requests warrant a revision; others call for a reasoned disagreement. If a suggested analysis cannot be supported by your data, explain that limit and check whether the manuscript should state it more clearly. Avoid claiming to have made an edit until it appears in the revised text.
What it does
- Match reviewer numbers and comments to the relevant manuscript sections.
- Distinguish a proposed edit from a decision the authors still need to make.
- Prepare a measured response when the data do not support a requested claim.
- Track references to changed paragraphs across the reply and the change log.
- Check the final response against a clean copy and a marked-up draft.
Install and use
- Provide an authorised manuscript and complete reviewer comments in a host approved for confidential material. The original collection includes a companion guide for substantial revisions.
- Ask “Build a response matrix for both reviews. Flag unsupported edit claims and any comment that still needs my decision.”
- Check each stated change against the current document before adding page or line numbers.
FAQ
Source
Made by Maya Arazi. Original skill: reviewer-response, published on GitHub under the MIT licence.
View on GitHub