Elicit Paper Screening Checklist from Study Notes (No Invented Citation Counts)

Prompt

Act as an Elicit paper-screening coordinator who only uses pasted study notes. You compile a paper-screening checklist the notes already support. You do not invent citation counts, effect sizes, journal ranks, or p-value claims. This is not a live Elicit sync, not Connected Papers graph merge, and not research or clinical advice.
You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs.

Inputs:
- Study notes I lock (paper stubs, inclusion cues, outcome fragments): [StudyNotes]
- Elicit version or workspace notes I lock: [Version]
- Study or review label I may quote (or UNKNOWN): [StudyLabel]
- Paper labels already present (or UNKNOWN): [PaperLabels]
- Inclusion criteria cues already present (or UNKNOWN): [InclusionCues]
- Outcome cues already present (or UNKNOWN): [OutcomeCues]
- Screening stage cues already present (or UNKNOWN): [ScreeningStageCues]
- Words I must not use: [Banned]
- What I must never invent (citation counts, effect sizes, journal ranks, p-value claims): [Never]
- Output format: [Format]
- Language: [Lang]

Generate:
1. Honesty ledger: StudyNotes nouns, Version, StudyLabel, PaperLabels, InclusionCues, OutcomeCues, ScreeningStageCues, Lang. Banner: not research advice; not a live Elicit sync. Forbidden: invented citation counts, effect sizes, journal ranks, p-value claims.
2. Paper-screening checklist: one checkbox row per PaperLabels entry. Attach only InclusionCues and OutcomeCues named beside that paper in StudyNotes. Missing cue write NOT IN INPUTS.
3. Screening-stage sketch: for each ScreeningStageCues entry, list papers that name it. Do not invent a 412 citation claim if absent.
4. Inclusion caution block: quote InclusionCues only. Meta-analysis packs not in StudyNotes stay NOT IN INPUTS.
5. Refuse list: inventing 412 citation counts, inventing 0.42 effect sizes, inventing journal rank #1, inventing p-value claims.
6. Compliance pass: quote Banned and Never hits. Cut them. Print paper and inclusion counts from StudyNotes only. Format as Format.

Constraints:
- Paper-screening checklist from StudyNotes only. No invented citation counts.
- Honor Version. No emojis. Not a live Elicit dashboard. Not research advice.

Instructions

Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.

Generated Output

This image was generated using the prompt above.

Elicit Paper Screening Checklist from Study Notes (No Invented Citation Counts) - Result

Examples

Example Input

StudyNotes: paper label Harbor Sleep RCT as pasted inclusion Adults as pasted outcome Sleep Latency as pasted; paper label Quay Nap Cohort as pasted inclusion Shift Workers as pasted. Citation invent NONE. Effect invent NONE.
Version: Elicit as pasted (do not invent unreleased AI meta coach)
StudyLabel: Cedar Pier sleep review as pasted
PaperLabels: Harbor Sleep RCT as pasted; Quay Nap Cohort as pasted. Background Protocol UNKNOWN.
InclusionCues: inclusion Adults as pasted for Harbor Sleep RCT; inclusion Shift Workers as pasted for Quay Nap Cohort. Pediatric pack UNKNOWN.
OutcomeCues: outcome Sleep Latency as pasted for Harbor Sleep RCT; Quay Nap Cohort outcome NOT IN INPUTS.
ScreeningStageCues: stage Title Abstract as pasted for Harbor Sleep RCT; Quay Nap Cohort stage NOT IN INPUTS.
Banned: 412 citations, 0.42 effect size, guaranteed journal rank #1
Never: invent citation counts, invent effect sizes, invent journal ranks, invent p-value claims
Format: ledger + paper-screening checklist + screening-stage sketch + inclusion caution + refuse + compliance
Lang: English

Example Output

1. Ledger. StudyNotes: paper label Harbor Sleep RCT + inclusion Adults + outcome Sleep Latency; paper label Quay Nap Cohort + inclusion Shift Workers. Citation invent NONE. Effect invent NONE. Version Elicit. StudyLabel Cedar Pier sleep review. PaperLabels Harbor Sleep RCT; Quay Nap Cohort. InclusionCues Adults for Harbor Sleep RCT; Shift Workers for Quay Nap Cohort. OutcomeCues Sleep Latency for Harbor Sleep RCT; Quay Nap Cohort outcome NOT IN INPUTS. ScreeningStageCues Title Abstract for Harbor Sleep RCT; Quay Nap Cohort stage NOT IN INPUTS. Lang English. Banner: not research advice; not a live Elicit sync. Forbidden: invented citation counts, effect sizes, journal ranks, p-value claims, 412 citations, 0.42 effect size, guaranteed journal rank #1.

2. Paper-screening checklist.
[ ] Harbor Sleep RCT | inclusion Adults | outcome Sleep Latency as pasted.
[ ] Quay Nap Cohort | inclusion Shift Workers | outcome NOT IN INPUTS.
Background Protocol not attached. Pediatric pack not added.

3. Screening-stage sketch.
stage Title Abstract | paper Harbor Sleep RCT as pasted.
Quay Nap Cohort stage | NOT IN INPUTS.
Citation counts NOT IN INPUTS so do not invent 412 citations. Second Title Abstract cue not invented.

4. Inclusion caution. inclusion Adults as pasted for Harbor Sleep RCT; inclusion Shift Workers as pasted for Quay Nap Cohort. Meta-analysis pack NOT IN INPUTS. Do not invent p-value packs.

5. Refuse. 412 citation counts: refused. 0.42 effect sizes: refused. journal rank #1: refused. p-value claims: refused. Unreleased AI meta coach: refused.

6. Compliance. Banned hits none. Papers 2. Inclusions named 2. Format ledger+paper-screening checklist+screening-stage sketch+inclusion caution+refuse+compliance. Gaps: Quay Nap Cohort outcome, Quay Nap Cohort stage, Background Protocol, Pediatric pack, citation counts.

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.