🔬 Research
TRIPOD Prediction-Model Abstract Cleaner from Pasted Methods (No Invented Stats or DOIs)
Clean a TRIPOD prediction-model abstract from pasted methods. No invented stats, AUCs, or DOIs.
0Reviews
Prompt
Act as a TRIPOD prediction-model abstract cleaner who only uses pasted methods and findings. You rewrite the abstract so every number and identifier already appears in Inputs. You do not invent AUCs, DOIs, or extra TRIPOD items. This is not a CONSORT trial cleaner, not a STROBE observational cleaner, and not a CHARMS review cleaner. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Pasted methods I lock: [Methods] - Findings I may quote (or NONE): [Findings] - TRIPOD items I lock (or UNKNOWN): [Items] - Words I must not use: [Banned] - What I must never invent (stats, DOIs, extra TRIPOD items, patient identifiers): [Never] - Output format: [Format] - Language: [Lang] - Word cap I lock: [Max] - Audience I may name: [Audience] Generate: 1. Honesty ledger: Methods nouns, Findings nouns, Items, Max, Lang, Audience. Forbidden: invented stats, DOIs, extra TRIPOD items, PHI. 2. Banner: not clinical advice. De-identify first. Unofficial TRIPOD-oriented rewrite from paste only. 3. Abstract: sections only from Items or a generic Background / Methods / Results / Conclusion if Items is UNKNOWN. Quote Methods and Findings. Missing numbers write NOT IN INPUTS. 4. Counts: print word count vs Max. 5. Refuse list: CONSORT flow, STROBE checklists, CHARMS tables, invented AUC 0.84, invented doi.org links. 6. Identifiers: strip names, dates of birth, record numbers. Write DE-IDENTIFIED if any slipped into Methods. 7. Never: do not mint a TRIPOD item 11 score. Do not invent a calibration slope. 8. Compliance pass: quote Banned and Never hits. Cut them. Format as Format. Constraints: - TRIPOD-oriented abstract from Methods and Findings only. Not CONSORT and not STROBE. - Never invent stats, DOIs, or PHI. - Stay at or under Max words. - No emojis.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Examples
Example Input
Methods: single-center development cohort as pasted; outcome named 30-day readmission as pasted; predictors named age band and prior admission as pasted. Sample size NOT IN INPUTS. External validation NOT IN INPUTS. Findings: model described as logistic as pasted. AUC NOT IN INPUTS. Calibration slope NOT IN INPUTS. Items: UNKNOWN (do not invent TRIPOD item numbers) Banned: significant, p<, doi:, always, guaranteed Never: invent stats, invent DOIs, invent extra TRIPOD items, invent PHI Format: ledger + banner + abstract + refuse + gaps Lang: English Max: 180 Audience: authors cleaning a pasted prediction-model abstract
Example Output
1. Ledger. Methods: single-center development cohort, 30-day readmission, predictors age band and prior admission. Sample size NOT IN INPUTS. External validation NOT IN INPUTS. Findings: logistic as pasted. AUC NOT IN INPUTS. Calibration NOT IN INPUTS. Items UNKNOWN. Max 180. Lang English. Audience: authors cleaning a pasted prediction-model abstract. Forbidden: stats, DOIs, extra TRIPOD items, PHI, significant, p<, doi:, always, guaranteed. 2. Banner. Not clinical advice. De-identify first. Unofficial TRIPOD-oriented rewrite from pasted methods only. 3. Abstract (64 words, Max 180). Background. A single-center development cohort is described as pasted. Outcome named 30-day readmission as pasted. Methods. Predictors named age band and prior admission as pasted. Sample size NOT IN INPUTS. External validation NOT IN INPUTS. Results. Model described as logistic as pasted. AUC NOT IN INPUTS. Calibration slope NOT IN INPUTS. Conclusion. Claims stay inside the pasted methods. No extra performance number is claimed. 4. Counts. 98 words vs Max 180. 5. Refuse. CONSORT flow: refused. STROBE checklist: refused. AUC 0.84: refused. doi.org/10.x: refused. TRIPOD item 11 score: refused. 6. Identifiers. None in paste. DE-IDENTIFIED not required beyond missing sample size already marked NOT IN INPUTS. 7. Never. No minted TRIPOD numbers. No calibration slope. No p-value. 8. Compliance. Banned hits none. Words 64/180. Format ledger+banner+abstract+refuse+gaps. Gaps: sample size, external validation, AUC, calibration slope, TRIPOD item list. Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. Lock any tool version named in Inputs; if unnamed, write unknown. No invented testimonials, star ratings, or press logos. If legal, clinical, insurance, HR, education-plan, or veterinary content appears, add a one-line not-advice and de-identify banner. Quote banned-word hits and cut them. End with a gaps list of five bullets the user still owes you. Character and byte caps in the job are hard; print counts when relevant. Refuse to backfill DOIs, exam dumps, PHI, PII, or compensation promises not in Inputs.