Paper notes — structured reading notes for research papers

What you get

A skill that turns the text of a research paper — full text, or abstract plus the sections you have — into one-page reading notes for a fashion team: claim, method and baselines, data and exact metrics, results as printed with references, limits stated by the authors and limits inferred (marked), relevance to a concrete fashion task, quotes with section references, open questions. It summarises only what you gave it and lists what was not provided. Also compares two papers, explains one to a non-researcher and triages a list of abstracts. Runs in Claude, ChatGPT, Gemini and Claude Code.

Use it for

• Write one-page reading notes from a paper's text, with every number referenced and every inference marked
• Compare two papers and find out whether their results are comparable at all
• Explain a paper to a designer or merchandiser in 200 words, ending with what it does not show
• Triage a reading list of abstracts: read fully, skim or skip, and what each abstract hides

What's inside

• SKILL.md — the skill: procedure, note format, self-checks, when to ask
• PROMPTS.md — Notes · Compare · Explain · Triage · Check prompts
• templates/reading-notes.md — the fillable one-page note
• examples/reading-notes-example.md — the template filled once for a fictional paper
• platforms/chatgpt-instructions.md — Custom GPT / Project instructions
• platforms/gemini-gem.md — Gem instructions
• platforms/claude-code/.claude/skills/paper-notes/SKILL.md — Claude Code drop-in

Quick start

1. Unzip and open README.md.
2. Claude: Settings → Capabilities → Skills → upload SKILL.md. ChatGPT: paste platforms/chatgpt-instructions.md into a Project. Gemini: paste platforms/gemini-gem.md into a Gem. Claude Code: copy platforms/claude-code/.claude/ into your project.
3. Paste the Notes prompt from PROMPTS.md with the paper text (or attach the PDF) and one line on why you are reading.

Preview — one prompt from the kit

Explain this paper to a colleague who does not read research (a designer, a merchandiser, a product manager), in at most 200 words, using ONLY the text below.

{{paper_text}} Their job: {{our_work}}

Rules: no jargon without a plain-word gloss; every number keeps its metric ("x % on the y test set"); say what was tested and on what data; one line on what would have to be true for it to matter to them; end with a line starting "What this does not show:". If a fact is not in the text, leave it out — do not add background from memory.