research-paper-writer
Write academic papers with AI assistance — outlining, drafting sections, tightening arguments, managing citations, and responding to reviewers. Use when turning research into a publishable manuscript.
- License 许可
- MIT
- 最近更新
- 2026-11-07
- Stars
- ★ 0
- Forks
- 1
01 前言 Front Matter
- name
- research-paper-writer
- category
- ai-research
- description
- Write academic papers with AI assistance — outlining, drafting sections, tightening arguments, managing citations, and responding to reviewers. Use when turning research into a publishable manuscript.
03 SKILL.md 正文
Research Paper Writer
Writing a paper is engineering an argument: the claim, the evidence, and the story that connects them. AI assistance helps at every stage — but the researcher owns the argument, the evidence, and the honesty.
Overview
Work section by section, not front to back. Start with the figures and results — the paper's skeleton. Write the methods from your lab notes while they're fresh. Draft the introduction last, when you know what story the results actually tell. Use AI for outlining, drafting, tightening prose, and checking consistency — never for inventing results, citations, or related work you haven't read.
When to use
- Turning completed research into a conference or journal manuscript.
- Structuring a messy set of results into a coherent narrative.
- Revising: tightening prose, fixing flow, responding to reviewers.
- Writing related work honestly from papers you've actually read.
Core concepts
- Argument first: one central claim per paper, supported by 2–4 sub-claims each tied to evidence. If you can't state the claim in one sentence, you're not ready to write.
- Results-driven structure: figures and tables first; the text explains what the reader sees. Write captions that stand alone.
- Methods reproducibility: enough detail for an expert to replicate — parameters, data versions, code links. Written from notes, not memory.
- Honest related work: cite what you read, characterize it fairly, position your contribution precisely. No strawmen, no citation padding.
- Reviewer response: address every comment — agree and change, or disagree with evidence. Never ignore; never be dismissive.
- Integrity boundaries: AI drafts prose; humans verify every claim, number, and citation. Fabricated references are career-ending — check each one.
Practical workflow
- State the one-sentence claim and list the evidence for each sub-claim. Cut results that don't serve the claim (save them for the next paper).
- Build figures and tables; write standalone captions.
- Outline section by section: the point of each paragraph in one line. Get feedback on the outline before drafting.
- Draft methods from lab notes; results describing the figures; discussion interpreting honestly — including what didn't work.
- Write the introduction last: the problem, why it matters, what you did, what you found — in that order.
- Revise in passes: argument → structure → clarity → precision → proofread. Use AI for tightening, then verify every edit preserved meaning.
Revision passes (one at a time):
1. ARGUMENT: does every section serve the central claim?
2. STRUCTURE: does the order build the case logically?
3. CLARITY: would a non-specialist follow the story?
4. PRECISION: are claims hedged correctly, numbers exact?
5. PROOFREAD: grammar, formatting, reference integrity
Common pitfalls
- Writing before the argument is clear: drafting to discover what you think. Outline first; the outline is where thinking happens.
- AI-invented citations: models fabricate plausible references. Verify every citation exists and says what you claim.
- Burying limitations: hiding weaknesses reviewers will find anyway. State them; frame what they imply for future work.
- Introduction first: writing the intro before results are final produces fiction. Results first, intro last.
- Ignoring the rebuttal: defensive or dismissive reviewer responses. Engage each point substantively — reviewers are doing you a favor.
- Salami slicing: splitting one contribution across papers. One paper, one clear contribution, fully evidenced.
04 同仓库更多 Skills
web-scrape
Turn web pages into LLM-ready markdown via the Firecrawl API (renders JS, anti-bot handling). Use when browser_open/fetch fails on JS-heavy pages, or when you need clean markdown/structured extraction from a site. Requires FIRECRAWL_API_KEY.
design-taste-frontend
tasteskill v2, the anti-slop frontend skill. Art direction for websites, distilled from a study of 247 curated gallery sites (One Page Love, Awwwards Sites of the Day). Deals researched directions (idea, composition, type, palette, photographs, section blueprints) from a library, enforces a strict no-slop house style with a real-browser inspector, and runs a look-and-fix loop on screenshots. Use this whenever you build, design, redesign, restyle, polish or review a landing page, marketing site, portfolio, product page, one-pager, hero or any public-facing web page, in any stack (HTML/CSS, React, Next.js, Vue, Svelte, Astro, Tailwind) — also when the user says a page looks generic, templated, "AI-made", sloppy, or asks to make it look designed, premium, award-level or less like every other site. Not for dashboards or app screens.
test-driven-development
Use when implementing any feature or bugfix, before writing implementation code
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
requesting-code-review
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).