Critique-and-revise loop: score a draft with a rubric and improve it in two rounds
Have the AI score your draft with a rubric, rewrite it, honestly re-score the new version and run a second round, with a change log and no invented facts. Designed for ChatGPT, Claude, Gemini and Microsoft Copilot. It includes 4 customizable variables and you can copy it for free in one click.
Intermediate · Text · Pending editorial review
Prompt
Act as a demanding editor who works in iterations: first critique with judgment, then revise, then re-assess your own work without self-congratulation.
Material:
- Draft:
"""
{{draft}}
"""
- Type of text and goal: {{goal}}
- Intended reader: {{reader}}
- Criteria I care about most: {{criteria}}
Process:
1. Round 1, critique: score the draft from 1 to 5 on clarity, structure, strength of argument, concision, tone fit for the reader and language correctness. Justify each score by quoting specific passages. Name the 3 problems that most hurt the goal.
2. Round 1, revision: rewrite the text to fix those 3 problems.
3. Round 2, critique: score the revised version with the same rubric, honestly (a score may not go up), and point out what still fails, including any new problems the revision introduced.
4. Round 2, revision: apply the remaining improvements. If every score is already 4 or higher, make only light edits and say so.
5. Wrap-up: a table showing how the scores changed, a log of the main changes and a list of things I need to decide or verify myself.
Constraints:
- Keep my voice, my stance and the meaning of the text; do not turn it into something generic.
- Do not add data, real examples or quotes that are not in the draft; if they are needed, use [FILL IN].
- Keep a similar length unless concision requires otherwise.
- If the goal or the reader is unclear, ask me before you start.
Format: headings per round, the full final version at the end and the score table.How to use it
Paste the full draft and state the goal in one concrete sentence (for example, 'get a meeting'). If round two still does not convince you, reply with the criterion that fails and ask for a third round focused only on it.
Usage example
Draft: 400-word proposal to sponsor an event · Goal: get a meeting with the brand · Reader: marketing manager · Criteria: clarity and persuasion.
Sample output
Assuming the draft as pasted (an email of about 200 words):
Round 1: critique
| Criterion | Score | Why |
|---|---|---|
| Clarity | 2 | The 8% increase only appears in paragraph three |
| Structure | 3 | Long background before the news |
| Argument | 2 | "Our costs have gone up" without saying which |
| Concision | 3 | Two sentences repeat the thank-you |
| Tone | 4 | Warm, right for a five-year relationship |
| Correctness | 4 | One typo in the greeting |
The 3 problems that hurt most: the number comes late, the justification is vague, and there's no next step.
Round 1: revision (opening)
Hi [NAME], I want to give you plenty of notice that starting [DATE], our rates will increase by 8%. Here's why: [FILL IN: specific costs]. Over these five years we've worked hard not to pass on…
Round 2: critique
Clarity rises to 4 and argument to 3: the concrete cost detail is still missing, and only you can supply it. New issue: the closing line sounds a little defensive.
Score changes
| Criterion | R0 | R1 | R2 | |
|---|---|---|---|---|
| Clarity | 2 | 4 | 4 | |
| Argument | 2 | 3 | 4 | … |
Key facts
- Category
- Writing & communication · Proofreading
- Output type
- Text
- Level
- Intermediate
- Designed for
- ChatGPT, Claude, Gemini and Microsoft Copilot
- Verified in
- Pending editorial verification
- Variables
- Draft, Type of text and goal, Reader and Priority criteria
- Language
- English
- Price
- Free, no sign-up
- Updated
- Nov 7, 2026
Frequently asked questions
Why ask for two rounds instead of one edit?
The first revision often fixes some problems and creates others. Re-scoring with the same rubric forces the model to check whether the text really improved.
Are the rubric scores reliable?
They are a guide to steer the revision, not an objective measure. The useful part is the justification with specific passages, which lets you decide whether you agree.
What is this prompt for?
Have the AI score your draft with a rubric, rewrite it, honestly re-score the new version and run a second round, with a change log and no invented facts.
Which AI tools does it work with?
It is designed for ChatGPT, Claude, Gemini and Microsoft Copilot. We have not verified it in these tools yet, so results may vary between them.
How do I customize the prompt?
Click “Customize” and fill in the variables (Draft, Type of text and goal, Reader and Priority criteria). The text updates instantly and you can then copy it in one click. Values are stored only in your browser.
Do I need experience to use it?
It is intermediate: it helps to know the topic so you can provide good context and review the result critically.
Is this prompt free?
Yes. Every HomeCasa Prompts prompt is free, with no sign-up or subscription. You can also use them in professional projects.
Can I trust the AI output?
Treat it as a high-quality draft: always check data, figures and important claims before publishing or making decisions.