AutoResearchClaw scientific-writing

Academic manuscript writing with IMRAD structure, citation formatting, and reporting guidelines. Use when drafting or revising research papers.

install
source · Clone the upstream repo
git clone https://github.com/aiming-lab/AutoResearchClaw
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/aiming-lab/AutoResearchClaw "$T" && mkdir -p ~/.claude/skills && cp -r "$T/researchclaw/skills/builtin/experiment/scientific-writing" ~/.claude/skills/aiming-lab-autoresearchclaw-scientific-writing-d74444 && rm -rf "$T"
manifest: researchclaw/skills/builtin/experiment/scientific-writing/SKILL.md
source content

Scientific Writing Best Practice

IMRAD Structure

  1. Abstract: State objective, methods, key results, and conclusion in 150-300 words
  2. Introduction: Move from broad context to specific gap to your contribution (funnel structure)
  3. Methods: Sufficient detail for replication; use past tense, passive voice
  4. Results: Present findings without interpretation; pair text with figures/tables
  5. Discussion: Interpret results, compare with literature, acknowledge limitations, state implications

Paragraph-Level Guidance

  1. Each paragraph should convey ONE main idea
  2. Open with a topic sentence; close with a transition to the next paragraph
  3. Write in full flowing prose — never submit bullet points as final manuscript text
  4. Use active voice for clarity: "We measured..." not "Measurements were taken..."
  5. Vary sentence length; aim for average 15-25 words per sentence

Citation Best Practices

  1. Cite primary sources over reviews when making specific claims
  2. Use citation styles consistently (APA, Vancouver, IEEE) per target journal
  3. Every factual claim needs a citation unless it is common knowledge in the field
  4. Avoid citation strings of 5+ references — select the most relevant 2-3
  5. Self-citations should be limited to genuinely relevant prior work

Common Writing Pitfalls

  1. Avoid hedge-stacking: "It might possibly suggest..." — choose one hedge
  2. Do not start sentences with "It is well known that" — cite or remove
  3. Distinguish "significant" (statistical) from "substantial" (practical)
  4. Ensure figures/tables are referenced in text BEFORE they appear
  5. Keep abbreviations to a minimum; define each on first use

Reporting Guidelines

  1. Randomized trials: follow CONSORT checklist
  2. Observational studies: follow STROBE checklist
  3. Systematic reviews: follow PRISMA checklist
  4. Diagnostic accuracy: follow STARD checklist
  5. Always check target journal's author guidelines for specific requirements

Revision Checklist

  1. Verify all figures/tables are cited in text and numbered sequentially
  2. Confirm reference list matches in-text citations exactly
  3. Check that abstract accurately reflects the final manuscript content
  4. Ensure methods section enables independent replication
  5. Read aloud to catch awkward phrasing and run-on sentences