scientific-paper-writer

Scientific Paper Writer

Section-by-section academic manuscript writing assistant for quantitative research. Version 1.0.0.

---

Overview

Scientific Paper Writer is a Claude plugin that guides researchers through writing a complete manuscript in expert writing order -- starting with Methods, then Results, Discussion, and frontmatter last. It supports 11 study types, verifies every citation against multiple databases, and exports to LaTeX or DOCX with venue-specific formatting.

Key principles:

• No fabrication -- all statistics are extracted directly from your analysis outputs, figures, and tables. Values that cannot be verified are flagged with [PLEASE VERIFY].
• Guideline-aligned -- every section follows the appropriate reporting guideline (PRISMA 2020, STROBE, CONSORT, TRIPOD+AI, STARD 2015, CHEERS 2022, COREQ/SRQR, PRISMA-NMA, PRISMA-IPD).
• Expert writing order -- Methods first, Abstract and Title last, following the order that experienced researchers use.
• Citation verification -- every external citation is verified against PubMed, CrossRef, Semantic Scholar, and bioRxiv/medRxiv before inclusion. Unverifiable citations are always flagged, never silently included.

---

Installation

Option 1: Install from GitHub (recommended)

claude plugin install github:olaTechie/scientific-paper-writer

Option 2: Install from a release zip

Download the zip from the Releases page, then:

claude plugin install ./scientific-paper-writer-v1.0.0.zip

Option 3: Install from local clone

git clone https://github.com/olaTechie/scientific-paper-writer.git
claude plugin install ./scientific-paper-writer

Verify installation

Type / in a Claude session. You should see all 8 commands listed below.

---

Quick Start

1. Create a PAPER_CONTEXT.md file in your working directory (see template below)
2. Follow the expert writing order:

/write-methods           Write Methods aligned with the reporting guideline
/write-results           Write Results from your analysis outputs
/write-discussion        Write Discussion with verified literature citations
/write-frontmatter       Write Abstract, Introduction, and Title
/write-captions          Write figure and table captions
/write-references        Build and verify the reference list
/generate-figures        Generate publication-quality figures (R/Python)
/export-manuscript       Export to LaTeX or DOCX

---

Commands

Command	Description	Example
/write-methods [study-type]	Write the Methods section aligned with the reporting guideline for your study type	/write-methods systematic-review
/write-results [study-type]	Write the Results section, extracting all numbers from source materials	/write-results rct
/write-discussion [section]	Write the Discussion with literature comparison and citation verification	/write-discussion main-findings
/write-frontmatter [section]	Write Abstract, Introduction, and Title after the body is complete	/write-frontmatter abstract
/write-captions [folder-path]	Generate publication-ready captions for figures and tables	/write-captions ./figures/
/write-references [style]	Verify citations and format the reference list	/write-references vancouver
/generate-figures [type]	Generate figures using R or Python scripts	/generate-figures forest-plot
/export-manuscript [format] [venue]	Export the manuscript to LaTeX and/or DOCX	/export-manuscript latex jama

---

Supported Study Types

Identifier	Reporting Guideline	Key Features
systematic-review	PRISMA 2020	PRISMA flow, risk of bias, pooled estimates, heterogeneity, subgroup/sensitivity analyses
network-meta-analysis	PRISMA-NMA	Network geometry, inconsistency testing, SUCRA/P-score rankings, league tables
ipd-meta-analysis	PRISMA-IPD	One-stage/two-stage models, participant-level subgroups, missing data handling
dose-response	PRISMA 2020 + dose-response extensions	Restricted cubic splines, non-linearity testing, threshold detection
bayesian-meta-analysis	PRISMA 2020 + Bayesian extensions	Prior specification, MCMC diagnostics, posterior summaries, model comparison
rct	CONSORT 2010	CONSORT flow, ITT/PP analyses, primary/secondary outcomes, adverse events
observational	STROBE	Cohort/case-control/cross-sectional, crude + adjusted estimates, bias assessment
economic-evaluation	CHEERS 2022	Cost-effectiveness analysis, ICER, sensitivity analyses, CEAC, budget impact
prediction-model	TRIPOD+AI	Model development, calibration, discrimination, internal/external validation