Verified against ChatGPT · 2026-08-13
Get essay feedback ranked by what would actually raise your grade, not a scattershot list of every issue found
Scores your draft against the actual rubric criteria, then ranks feedback by realistic point impact so you know what to fix first with limited revision time, instead of an undifferentiated wall of comments.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Give me feedback on this essay draft, but I need it prioritized by what will actually move my grade, not a flat list of every issue you notice — I have limited time before this is due and need to know what to fix first. ESSAY DRAFT [Full text of a 5-paragraph essay on the causes of the fall of the Roman Republic] RUBRIC OR GRADING CRITERIA Thesis clarity (20%), use of evidence (30%), organization (20%), analysis depth (20%), mechanics (10%) ASSIGNMENT PROMPT Argue whether economic inequality or institutional breakdown was the more significant cause of the Republic's fall, using at least two primary sources TIME I HAVE TO REVISE About 90 minutes tonight before it's due tomorrow morning First, go criterion by criterion through Thesis clarity (20%), use of evidence (30%), organization (20%), analysis depth (20%), mechanics (10%) and give an honest current-draft assessment against each one — not a vague impression, but which specific part of the essay is or isn't meeting that criterion, quoting or referencing the actual passage. Then rank the issues you found by realistic point impact: which fix would move the score the most for the effort it takes, given About 90 minutes tonight before it's due tomorrow morning. A large structural issue (the thesis doesn't actually match what the body paragraphs argue) that's fixable in the time available should outrank a dozen small wording issues, even though the wording issues are individually easier to explain. For each of the top 3 priority issues, give a specific instruction for the fix — not "strengthen your thesis" but the actual sentence-level or paragraph-level change to make, tied to the actual text in [Full text of a 5-paragraph essay on the causes of the fall of the Roman Republic]. Explicitly flag if the essay doesn't actually answer Argue whether economic inequality or institutional breakdown was the more significant cause of the Republic's fall, using at least two primary sources as asked, since that's a category of problem no amount of line-editing fixes and it needs to be caught before anything else. WHAT NOT TO DO Do not rewrite the essay for me or produce a corrected version — that's not what's being asked, and it would replace my thinking rather than improve my draft. Do not comment on every small grammar issue in prose; if there are recurring small issues, name the pattern once rather than flagging each instance. OUTPUT FORMAT 1. Criterion-by-criterion assessment (rubric item -> current state -> met/partially met/not met). 2. Whether the essay actually answers the assignment prompt as asked. 3. Top 3 priority fixes, ranked, each with a specific actionable instruction. 4. One line on what NOT to spend revision time on given About 90 minutes tonight before it's due tomorrow morning.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
An unconstrained feedback request produces a flat, undifferentiated list where a missing comma and a thesis that doesn't match the argument get roughly equal billing, because GPT-5.1 without a ranking instruction tends to surface issues in the order it notices them while reading, not in the order that matters for the grade — explicitly requiring a rubric-anchored, point-impact ranking forces the model to weigh severity against the actual grading weights rather than treating every observation as equally worth mentioning. Requiring the criterion-by-criterion pass to reference actual passages from {{essay_text}} rather than give an impressionistic summary keeps the assessment grounded and falsifiable — a vague "your evidence use is a bit thin" is not actionable, but pointing to the specific paragraph where a claim has no citation is. Explicitly checking whether the essay answers {{assignment_prompt}} as asked, as a distinct and prioritized first check, matters because this is the single highest-impact category of essay problem and the easiest for both a rushed student and a surface-reading model to miss — an essay that's well-written but argues a different question than the one assigned will lose points no amount of sentence-level polish recovers, and it needs to surface before line-level feedback, not buried inside a general comment. Weighting fixes by effort-to-impact given the actual time budget in {{time_available}} is what turns generic feedback into a usable revision plan under real constraints — the same feedback that's correct in the abstract is useless if it recommends restructuring three paragraphs when the student has ninety minutes, and asking the model to reason about realistic revision scope rather than idealized thoroughness is what keeps the top-3 list actually followable before the deadline.
What you get back
Thesis clarity: PARTIALLY MET — your intro thesis states inequality was 'the primary cause,' but paragraphs 3 and 4 spend more analytical depth on institutional breakdown (the Gracchi reforms, Senate gridlock) than on inequality itself, creating a mismatch a grader will notice. Priority fix #1 (highest impact, ~20 min): either revise the thesis to reflect that institutional breakdown gets the stronger argument in your body paragraphs, or add a paragraph of evidence that actually supports the inequality claim at the same depth...
Verified against
ChatGPT GPT-5.1 · 2026-08-13
Changelog
- 2026-08-13 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
