paper-writer
VerifiedMedical/scientific paper writing workflow skill. Manages the full pipeline from literature search to submission-ready manuscript. Creates and manages a project directory with IMRAD-format section files, literature matrix, reference management, and quality checklists. Supports both English and Japanese papers. Triggers: 'write paper', 'paper-write', 'start manuscript', '論文を書く', '論文執筆', '論文プロジェクト', 'manuscript', 'research paper', '原稿作成'.
templates/sr-prospero.md
View on GitHub# PROSPERO Registration Template
> PROSPERO is the international prospective register of systematic reviews.
> Register your protocol BEFORE starting screening.
> Registration URL: https://www.crd.york.ac.uk/prospero/
> Average processing time: 2-4 weeks.
---
## Instructions
1. Complete all fields below before submitting to PROSPERO.
2. Fields marked **(required)** must be filled; others are strongly recommended.
3. Prepare your text offline using this template, then copy into the PROSPERO web form.
4. Character limits are noted where applicable.
5. After registration, you will receive a CRD number (e.g., CRD42026XXXXXX).
---
## 1. Review Title **(required)**
**Field:** Full title of the systematic review.
**Tips:**
- Include study design (systematic review, meta-analysis)
- Use PICO/PECO elements in the title
- Avoid abbreviations
**Common mistakes:**
- Title too vague (e.g., "Review of AI in healthcare")
- Missing "systematic review" in title
**Example:**
> Large Language Models for Patient Education Material Generation in Chronic Disease Management: A Systematic Review and Meta-Analysis
**Your entry:**
{{TITLE}}
---
## 2. Original Language Title
**Field:** Title in the original language if not English.
**Your entry:**
{{ORIGINAL_TITLE}}
---
## 3. Anticipated or Actual Start Date **(required)**
**Field:** The date the review started or is expected to start (DD/MM/YYYY).
**Tips:**
- This is when you begin searching, not when you register
- Must be a future date or very recent past
- PROSPERO will reject if screening has already been completed
**Common mistakes:**
- Setting a date in the distant past (raises red flags)
- Submitting after screening is complete (PROSPERO may reject)
**Example:**
> 01/03/2026
**Your entry:**
{{START_DATE}}
---
## 4. Anticipated Completion Date **(required)**
**Field:** The date by which the review is expected to be completed (DD/MM/YYYY).
**Tips:**
- Allow 6-12 months for a thorough SR
- Be realistic; you can update later
**Your entry:**
{{COMPLETION_DATE}}
---
## 5. Stage of Review at Submission **(required)**
**Field:** Select the current stage.
**Options:**
- [ ] Preliminary searches
- [ ] Piloting of the study selection process
- [ ] Formal screening against eligibility criteria
- [ ] Data extraction
- [ ] Risk of bias assessment
- [ ] Data analysis
**Tips:**
- Ideally submit at "Preliminary searches" stage
- PROSPERO accepts up to "Data extraction" stage
- Registration after data extraction has begun will NOT be accepted
**Your entry:**
{{STAGE}}
---
## 6. Named Contact **(required)**
**Field:** Name of the guarantor for the review.
**Your entry:**
{{CONTACT_NAME}}
---
## 7. Named Contact Email **(required)**
**Field:** Institutional email address of the named contact.
**Your entry:**
{{CONTACT_EMAIL}}
---
## 8. Named Contact Address **(required)**
**Field:** Full institutional postal address.
**Your entry:**
{{CONTACT_ADDRESS}}
---
## 9. Named Contact ORCID
**Field:** ORCID of named contact.
**Your entry:**
{{ORCID}}
---
## 10. Review Team Members and Affiliations **(required)**
**Field:** List all team members with affiliations and roles.
**Tips:**
- Include all authors who will contribute
- Specify roles (e.g., screening, extraction, analysis)
- Include a medical librarian if involved in search strategy
**Example:**
> 1. Dr. Ken Okamoto, Department of Medicine, University of Tokyo - Lead reviewer, screening, data extraction, analysis
> 2. Dr. Jane Smith, School of Public Health, University of Oxford - Screening, risk of bias assessment
> 3. Mr. John Doe, Medical Library, University of Tokyo - Search strategy development
> 4. Prof. Sarah Lee, Department of Biostatistics, Kyoto University - Statistical analysis, meta-analysis
**Your entry:**
{{TEAM_MEMBERS}}
---
## 11. Funding Sources/Sponsors **(required)**
**Field:** Details of funding or statement of no funding.
**Tips:**
- Include grant numbers
- State "No external funding" if self-funded
- Disclose industry funding
**Example:**
> This review is supported by JSPS KAKENHI Grant Number JP26XXXXXXX. The funder has no role in the design, conduct, or reporting of this review.
**Your entry:**
{{FUNDING}}
---
## 12. Conflicts of Interest **(required)**
**Field:** Declare any conflicts of interest for all team members.
**Tips:**
- "None declared" is acceptable if true
- Disclose any relationships with companies whose products are reviewed
**Example:**
> The authors declare no conflicts of interest relevant to this systematic review.
**Your entry:**
{{COI}}
---
## 13. Collaborators
**Field:** Additional contributors not listed as review team members.
**Your entry:**
{{COLLABORATORS}}
---
## 14. Review Question **(required)**
**Field:** State the review question clearly using a structured format.
**Tips:**
- Use PICO (intervention reviews) or PECO (exposure reviews)
- Be specific and focused
- One primary question; secondary questions listed separately
**Common mistakes:**
- Too broad (e.g., "What is the effect of AI?")
- Combining multiple unrelated questions
- Not using a structured framework
**Example:**
> In adult patients with type 2 diabetes (P), does AI-generated patient education material (I), compared to clinician-written material (C), improve health literacy, self-management behaviors, and glycemic control (O)?
**Your entry:**
{{REVIEW_QUESTION}}
---
## 15. Searches **(required)**
**Field:** List databases and other sources to be searched.
**Tips:**
- Minimum 2 databases (MEDLINE + one other)
- Include grey literature sources
- Name specific platforms (e.g., MEDLINE via PubMed, not just "MEDLINE")
- Mention hand-searching, reference checking, trial registries
**Common mistakes:**
- Only searching one database
- Not specifying the platform
- Omitting grey literature
**Example:**
> The following electronic databases will be searched from inception to the search date: MEDLINE (via PubMed), Embase (via Elsevier), Cochrane Central Register of Controlled Trials (CENTRAL), CINAHL (via EBSCOhost), and PsycINFO (via APA PsycNet).
>
> Grey literature will be searched via Google Scholar (first 200 results), ProQuest Dissertations & Theses, and conference proceedings from AMIA and MedInfo (2020-2026).
>
> Trial registries: ClinicalTrials.gov and WHO ICTRP.
>
> Reference lists of all included studies and relevant systematic reviews will be hand-searched.
**Your entry:**
{{SEARCHES}}
---
## 16. URL for Published Protocol
**Field:** Link to any published protocol (e.g., on OSF, journal).
**Your entry:**
{{PROTOCOL_URL}}
---
## 17. Condition or Domain Being Studied **(required)**
**Field:** The health condition, disease, or topic area.
**Example:**
> Patient education and health literacy in chronic disease management, specifically examining the role of artificial intelligence-generated educational materials.
**Your entry:**
{{CONDITION}}
---
## 18. Participants / Population **(required)**
**Field:** Describe the population with inclusion and exclusion criteria.
**Tips:**
- Be specific about age, condition, setting
- State exclusions clearly
**Example:**
> Adults (≥18 years) diagnosed with type 2 diabetes mellitus, managed in any healthcare setting (primary care, secondary care, community). Exclusions: gestational diabetes, type 1 diabetes, pediatric populations, animal studies.
**Your entry:**
{{POPULATION}}
---
## 19. Intervention(s), Exposure(s) **(required)**
**Field:** Describe the intervention or exposure of interest.
**Tips:**
- Be specific about type, dose, frequency, duration
- Include variations that will be accepted
**Example:**
> Patient education materials generated or substantially assisted by large language models (e.g., GPT-4, Claude, Gemini, LLaMA) or other AI/NLP systems. This includes fully AI-generated text, AI-assisted drafting with clinician review, and AI-translated/adapted materials. Materials may be delivered in any format (written, audio, video with AI-generated scripts).
**Your entry:**
{{INTERVENTION}}
---
## 20. Comparator(s) / Control **(required)**
**Field:** Describe the comparison group(s).
**Example:**
> Clinician-authored patient education materials, standard care patient education, commercially available patient education resources, or no patient education intervention.
**Your entry:**
{{COMPARATOR}}
---
## 21. Types of Study to Be Included **(required)**
**Field:** Study designs eligible for inclusion.
**Tips:**
- Specify designs clearly
- State any design hierarchy
**Common mistakes:**
- Being too restrictive (only RCTs) when evidence is sparse
- Being too inclusive without justification
**Example:**
> Randomized controlled trials (RCTs), quasi-experimental studies, controlled before-after studies, and prospective cohort studies with a comparison group. Case reports, case series, editorials, commentaries, conference abstracts without full text, and narrative reviews will be excluded.
**Your entry:**
{{STUDY_TYPES}}
---
## 22. Main Outcome(s) **(required)**
**Field:** Primary outcomes with measurement details.
**Tips:**
- Limit to 1-3 primary outcomes
- Specify how each will be measured
- Define minimum clinically important difference if known
**Example:**
> 1. Health literacy: Measured by validated instruments (e.g., NVS, REALM, HLS-EU-Q47) at any time point post-intervention.
> 2. Accuracy of information: Assessed by clinician review or validated accuracy scoring tools.
**Measures of effect:**
> Risk ratios (RR) with 95% confidence intervals for dichotomous outcomes. Mean differences (MD) or standardized mean differences (SMD) with 95% CI for continuous outcomes.
**Your entry:**
{{MAIN_OUTCOMES}}
---
## 23. Additional Outcome(s)
**Field:** Secondary outcomes.
**Example:**
> 1. Patient satisfaction with educational materials (any validated measure)
> 2. Self-management behaviors (e.g., medication adherence, dietary compliance)
> 3. Clinical outcomes (e.g., HbA1c change)
> 4. Readability scores (e.g., Flesch-Kincaid, SMOG)
> 5. Adverse events or harms (e.g., misinformation, anxiety)
**Your entry:**
{{ADDITIONAL_OUTCOMES}}
---
## 24. Data Extraction (Selection and Coding) **(required)**
**Field:** Describe the data extraction process.
**Example:**
> Two reviewers will independently extract data using a standardized, pilot-tested extraction form. Data items include: study characteristics (author, year, country, design, setting), participant characteristics (sample size, age, sex, condition), intervention details (AI model, prompt strategy, output format, clinician review process), comparator details, outcome measures and results (effect sizes with confidence intervals), and risk of bias ratings.
>
> Discrepancies will be resolved by discussion or consultation with a third reviewer. Study authors will be contacted for missing data with up to two email attempts over four weeks.
**Your entry:**
{{DATA_EXTRACTION}}
---
## 25. Risk of Bias (Quality) Assessment **(required)**
**Field:** Tools and process for assessing methodological quality.
**Tips:**
- Match the tool to study design
- Describe the process (independent, dual)
**Example:**
> Risk of bias will be assessed independently by two reviewers using the following tools:
> - RCTs: Cochrane Risk of Bias tool version 2 (RoB 2)
> - Non-randomized studies: Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I)
> - Cross-sectional studies: JBI Critical Appraisal Checklist for Analytical Cross-Sectional Studies
>
> Disagreements will be resolved by discussion or a third reviewer. Results will be presented in risk of bias summary tables and traffic light plots.
**Your entry:**
{{ROB_ASSESSMENT}}
---
## 26. Strategy for Data Synthesis **(required)**
**Field:** Describe how data will be synthesized.
**Tips:**
- Describe both narrative and quantitative approaches
- Specify meta-analysis model and software
- Define heterogeneity thresholds
- Describe subgroup and sensitivity analyses
**Example:**
> Studies will be grouped by outcome and synthesized narratively, describing direction and magnitude of effects.
>
> Where three or more studies report the same outcome with sufficient homogeneity, random-effects meta-analysis will be performed using the restricted maximum likelihood (REML) estimator in R (meta package, version 7.0). Heterogeneity will be assessed using the I² statistic and Cochran's Q test (significance at p < 0.10). Prediction intervals will be calculated.
>
> Pre-specified subgroup analyses: (1) AI model type, (2) clinician review vs. no review, (3) disease type, (4) study design.
>
> Sensitivity analyses: (1) excluding high risk of bias studies, (2) excluding studies with imputed data.
>
> Publication bias will be assessed using funnel plots and Egger's test when ≥10 studies are pooled.
>
> Certainty of evidence will be assessed using the GRADE approach.
**Your entry:**
{{SYNTHESIS_STRATEGY}}
---
## 27. Analysis of Subgroups or Subsets
**Field:** Planned subgroup or subset analyses.
**Your entry:**
{{SUBGROUP_ANALYSES}}
---
## 28. Type and Method of Review **(required)**
**Field:** Select the type of review.
**Options:**
- [x] Systematic review
- [ ] Meta-analysis (check if planned)
- [ ] Network meta-analysis
- [ ] Individual participant data meta-analysis
- [ ] Rapid review
- [ ] Scoping review
- [ ] Other
**Your entry:**
{{REVIEW_TYPE}}
---
## 29. Language
**Field:** Language restriction, if any.
**Example:**
> No language restrictions will be applied. Non-English articles will be translated using professional translation services or bilingual team members.
**Your entry:**
{{LANGUAGE}}
---
## 30. Country
**Field:** Country of the review team.
**Your entry:**
{{COUNTRY}}
---
## 31. Other Registration Details
**Field:** Any other protocol registration (e.g., OSF, journal protocol).
**Your entry:**
{{OTHER_REG}}
---
## 32. Reference and/or URL for Published Protocol
**Your entry:**
{{PROTOCOL_REF}}
---
## 33. Dissemination Plans
**Field:** Plans for publishing and sharing results.
**Example:**
> Results will be submitted to a peer-reviewed journal. Findings will be presented at relevant conferences (e.g., AMIA Annual Symposium). The completed review, including all data extraction forms and analysis code, will be deposited in an open-access repository (e.g., OSF, Zenodo).
**Your entry:**
{{DISSEMINATION}}
---
## 34. Keywords **(required)**
**Field:** 3-10 keywords for indexing.
**Example:**
> systematic review; large language models; patient education; health literacy; artificial intelligence; chronic disease
**Your entry:**
{{KEYWORDS}}
---
## 35. Details of Any Existing Review of the Same Topic
**Field:** Cite existing reviews and explain how this review differs.
**Tips:**
- Search PROSPERO and PubMed for existing reviews first
- Explain what is different about your review (scope, population, intervention, date)
**Example:**
> A scoping review by Chen et al. (2025) examined AI applications in patient education broadly but did not assess effectiveness or quality of evidence. Our review focuses specifically on LLM-generated materials, includes only comparative studies, and will assess certainty of evidence using GRADE.
**Your entry:**
{{EXISTING_REVIEWS}}
---
## 36. Current Review Status **(required)**
**Field:** Select the current status.
**Options:**
- [ ] Ongoing
- [ ] Completed but not published
- [ ] Published in full
- [ ] Discontinued
**Your entry:**
{{STATUS}}
---
## 37. Any Additional Information
**Field:** Any other relevant details.
**Your entry:**
{{ADDITIONAL_INFO}}
---
## Pre-Submission Checklist
- [ ] Title is descriptive and includes "systematic review"
- [ ] Review question uses PICO/PECO framework
- [ ] At least 2 databases listed with platforms
- [ ] Two independent reviewers specified for screening, extraction, and RoB
- [ ] RoB tool matched to study design
- [ ] Synthesis strategy clearly described
- [ ] GRADE assessment planned
- [ ] Subgroup analyses are pre-specified with rationale
- [ ] All team members listed with affiliations and roles
- [ ] COI and funding clearly stated
- [ ] Registration submitted BEFORE screening begins
- [ ] Existing reviews searched on PROSPERO
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |