AI Prompt for Grant Resubmission Using Reviewer Comments
This AI prompt for grant resubmission using reviewer comments turns the feedback on a declined proposal into a revision plan. It lists every criticism with the reviewer's words, maps each to the section it concerns, and separates presentation problems from weaknesses in design or evidence and from problems of fit that revision cannot fix. It shows where reviewers agreed and disagreed, ranks the changes, and outlines a point-by-point response if the funder allows one.
The prompt
You are helping me analyze the feedback on a declined grant proposal so I can decide whether and how to resubmit. Use only the material I paste below. Do not dismiss, soften or argue with any criticism. Do not guess at reviewers' identities, expertise or motives. Do not invent facts, data, citations, activities, partners, results, budgets or funder rules. Funding opportunity, briefly: [OPPORTUNITY SUMMARY] Our project, briefly: [PROJECT SUMMARY] Funder's resubmission rules: [PASTE RULES OR "NOT PROVIDED"] Reviewer comments, scores and any summary statement: [PASTE ALL FEEDBACK] Sections of the original proposal that the feedback refers to: [PASTE RELEVANT PROPOSAL SECTIONS] Step 1. List every distinct criticism. For each one: - quote the reviewer's wording - name the proposal section it concerns and quote the passage, if I pasted it - classify it as: (a) presentation or clarity, (b) design, evidence or methods, (c) feasibility or capacity, (d) budget, (e) fit with the funder's priorities or eligibility, or (f) other - say whether it looks fixable by rewriting, needs new evidence or a redesign, or may not be fixable by revision at all Step 2. Group the criticisms into themes. Show where reviewers agreed, where they disagreed and where a comment is ambiguous. Do not treat one reviewer's remark as the panel's view, and do not dismiss a point because only one reviewer raised it. Step 3. List what the reviewers said was strong, so I do not cut it in revision. Step 4. Give me a ranked revision plan. For each change: the section, what needs to change, the evidence or decision required from my team, and whether it needs new work or only new writing. Step 5. If the resubmission rules I pasted allow a response to reviewers, draft a neutral point-by-point outline: the concern, an acknowledgment, and a placeholder such as [DESCRIBE REVISION] for what we did. Do not claim changes we have not made. If the rules do not allow a response, or I did not provide the rules, do not draft one and tell me to check. Step 6. End with three lists: questions I must answer, facts I must verify, and criticisms that rewording alone cannot address. Do not tell me whether to resubmit. That is my decision.
Written for ChatGPT, Claude, Gemini and similar AI chat tools. Replace each bracketed placeholder with your own details before you send it. Results vary by tool and version, so always check the output.
What to fill in
- [PASTE RULES OR "NOT PROVIDED"]
- What the funder says about resubmitting: whether it is allowed, how many times, by when, and whether a response to the previous review is permitted or required. Check the current rules, since they may have changed since you first applied.
- [PASTE ALL FEEDBACK]
- Every comment and score you received, in full and unedited. Do not leave out the ones you disagree with. Before pasting, check that the funder's terms let you share the feedback with an AI tool.
- [PASTE RELEVANT PROPOSAL SECTIONS]
- The parts of the original proposal that reviewers commented on. Paste only what is needed, with confidential and personal information removed.
- [OPPORTUNITY SUMMARY]
- What the funder was looking for, in a few lines, so the analysis can tell weak writing from weak fit.
Why this prompt works
A rejection is hard to read clearly. People skim the comments, fix on the one that stung, and either dismiss the feedback or try to change everything. This prompt slows that down and turns the feedback into something you can work through.
- Every criticism is listed. Nothing is summarized away. Each point is quoted, so you can check the AI did not soften or sharpen it.
- Criticisms are sorted by kind. A clarity problem, a design problem and a fit problem call for different responses. Treating a design weakness as a writing problem is a common way for a resubmission to fail again.
- "May not be fixable" is an allowed answer. Sometimes the feedback says the project does not match what the funder wants. The prompt lets the AI say so instead of producing a revision plan for the wrong funder.
- Agreement and disagreement are shown. A concern raised by every reviewer is different from one raised by one. The prompt reports both without letting a single comment stand for the panel or be waved away.
- Strengths are protected. Step 3 records what reviewers liked, so revision does not remove it.
- No arguing and no mind reading. The AI is barred from disputing the criticism or speculating about who the reviewers were. Neither helps.
- A response outline with placeholders. The outline acknowledges each concern and leaves a blank for what you actually changed, so it cannot claim revisions that have not happened.
- The decision stays with you. The prompt declines to say whether you should resubmit.
Why it is appropriate to use
Use this after a decline, before you decide what to do next. The options are to revise and resubmit, redesign the project, take it to a different funder or set it aside. An organized reading of the feedback is what lets you choose among them.
Using AI to organize feedback is reasonable, with one important check first. Reviewer comments may be confidential. Some funders restrict how applicants and reviewers may share review materials, and some have specific rules about putting them into AI tools. Read the terms that came with your feedback and the funder's AI policy before you paste anything. If you are not permitted to share it, do this analysis by hand using the same steps.
Not every funder gives feedback, and resubmission rules vary widely. Some funders welcome a revised application, some limit how many times you can resubmit and some do not accept resubmissions at all. The funder's current rules govern.
Do not use the prompt to build a rebuttal, to find reasons the reviewers were wrong or to write around a weakness the panel identified. It cannot supply the new evidence, partners or results that a substantive criticism calls for. See grant submission, peer review and resubmission strategy for the wider approach.
How to check and revise the output
- Check every quotation and classification against the original feedback. Make sure nothing was left out.
- Take the criticisms of design and evidence to your project team. Decide honestly whether you can address them. If you cannot, rewording will not help.
- Look hard at anything classified as fit. If the project is not what this funder funds, a different funder may be the right next step.
- If the funder offers it, talk to the program officer about the feedback before you revise.
- Confirm the resubmission rules and deadline in the funder's current guidance.
- Make the changes first, then write the response. Fill each placeholder with what you actually did, in a neutral and specific tone.
- Have the revised proposal reviewed afresh by someone who has not seen the original.
Frequently asked questions
- Can I use AI to analyze grant reviewer comments?
- Only after checking the confidentiality terms attached to the feedback and the funder's AI policy. Where it is permitted, AI can help organize and classify the comments. You still need to verify its reading against the original feedback and make the decisions yourself.
- How should I respond to reviewer comments in a resubmission?
- Address each significant concern directly, describe the specific change you made and avoid defensiveness. Include a point-by-point response only if the funder permits one, and follow their format.
- What if the reviewers disagreed with each other?
- Look first for the concerns they shared. Then ask whether the disagreement comes from unclear writing or from a real difference of opinion. Do not assume one reviewer was wrong, and do not ignore a concern because only one person raised it.
- Should I resubmit a declined grant proposal?
- Consider it when the funder allows resubmission, the opportunity is still a good fit and you can credibly address the main concerns. If the feedback points to a mismatch with the funder's priorities, taking the project elsewhere may serve you better than rewriting it.