paper-reading

Convert research papers into structured decision-oriented notes.

27|4|Updated Mar 30, 2026
One-click install
npx skills add https://github.com/yishuai778/paper-reading --skill paper-reading-yishuai778
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: paper-reading
Source: https://github.com/yishuai778/paper-reading/tree/main
Command: npx skills add https://github.com/yishuai778/paper-reading --skill paper-reading-yishuai778

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

This Skill turns research papers into structured, decision-oriented notes that help users distinguish what the paper actually says from what can be reasonably inferred, enabling better trust, replication decisions, and practical use.

Core Features & Use Cases

  • Not a generic summarizer: it produces judgment-ready outputs that separate claims, evidence, and inferences.
  • Supports multiple reading modes (Quick Read, Deep Read, Comparison Read, Literature Seed Map, Reproduction Memo, HTML Reading Page) and can compare or summarize multiple papers.
  • Includes a design philosophy and workflow (frame the paper, parse the mechanism, judge the evidence, produce a decision-grade output) to guide decision-making and implementation planning.

Quick Start

Read this paper and produce a decision-grade analysis with problem framing, mechanism, and actionable takeaways.

Frequently Asked Questions about paper-reading

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I turn research papers into structured notes for trust and reuse decisions?

To turn research papers into structured notes, this skill separates claims, evidence, and inferences to produce decision-grade outputs that support judgments about trust, reuse, or reproduction. It frames the paper, parses the mechanism, and judges the evidence.

Can I compare multiple papers to build a literature seed map?

Yes, you can compare multiple papers using the Comparison Read and Literature Seed Map workflows. These modes analyze several papers together to help you evaluate evidence across studies and map related research domains.

What is the best way to evaluate evidence in a research paper?

The best way to evaluate evidence is to use a structured workflow that distinguishes what the paper actually says from what can be reasonably inferred. This skill frames the problem, parses the mechanism, and judges the evidence to guide implementation planning.

Does this skill generate an HTML reading page for papers?

Yes, the skill includes an HTML Reading Page workflow that converts research papers into structured, decision-oriented notes. It produces an HTML format output to facilitate reading and evaluating the paper's claims and evidence.

How do I create a reproduction memo from a research paper?

You can create a reproduction memo using the dedicated workflow that converts the paper's mechanism and evidence into a judgment-ready output. This structured note specifically supports decisions about whether to replicate or reproduce the study.

How do I do a quick read of a paper to extract actionable takeaways?

Use the Quick Read mode to rapidly process the paper and produce a decision-grade analysis. This workflow extracts problem framing, mechanism details, and actionable takeaways without deep parsing.