← ClaudeAtlas

scientific-writinglisted

Draft, review, and iteratively revise scientific manuscripts with a provider-agnostic multi-agent workflow for Codex and Claude Code. Use for new manuscripts, section rewrites, rebuttals, response letters, manuscript QA, or sentence-level writing reviews when claims must stay grounded in supplied artifacts.
fmschulz/omics-skills · ★ 3 · AI & Automation · score 67
Install: claude install-skill fmschulz/omics-skills
# Scientific Writing Use this as the single scientific-writing skill for both Codex and Claude Code. It covers two related modes of work: - **Drafting and revision** — produce a manuscript or rebuttal from authoritative artifacts and iterate it through review-revise loops. - **Sentence-level writing review** — audit existing prose for clarity, voice, sentence architecture, terminology consistency, and numerical integrity, without changing the underlying science. The two modes share the same role set and the same hard rules. Pick the entry point that matches the request; the workflow then routes accordingly. ## State Assumptions First Before drafting or reviewing, state: - manuscript mode: new draft, revision, rebuttal, review response, section rewrite, or writing review - review mode when this is a writing review: `full-review`, `section-review`, `targeted`, or `interactive` (see below) - target venue or formatting target when known - authoritative artifacts available - missing artifacts that block specific claims If a required input is missing, name the gap and keep the prose conservative. ## Review Modes When the request is a writing review of existing prose, choose one mode: | Mode | Trigger | Behavior | |---|---|---| | `full-review` | "review my manuscript", "writing review" | Run all five prose-quality audits on the whole document and emit a structured report. Default when ambiguous. | | `section-review` | "review the Introduction", "check the Discussion" | Run