Evaluating Paper Relevance
Two-stage paper screening - abstract scoring then deep dive for specific data extraction
npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill evaluating-paper-relevance --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Two-stage screening process: quick abstract scoring followed by deep dive into promising papers.
How it works
The skill defines a staged workflow:
- Stage 1: Abstract Screening (Fast)
- Score papers on a 0-10 scale using keywords match (0-3), data type match (0-4), and specificity (0-3).
- Apply decision rules: Score < 5 skip, Score 5-6 note as possibly relevant, Score ≥ 7 proceed to Stage 2.
- Require reporting to the user for every paper with explicit messages showing the abstract score and the next action (fetching full text or skipping).
- Stage 2: Deep Dive (Thorough)
- If medicinal chemistry is involved, use
skills/research/checking-chemblto check for ChEMBL data and annotate ChEMBL IDs and data presence. - Fetch full text in order: PubMed Central (free), DOI resolution, Unpaywall (MANDATORY if paywalled), Preprints (bioRxiv/arXiv).
- If full text is unavailable, note the paywall status and continue with abstract-only evaluation.
- Scan for content in Methods, Results, Tables/Figures, and Supplementary Information to extract domain-specific data (e.g., IC50, expression levels, algorithms, datasets).
- Create structured extractions and update tracking files (papers-reviewed.json and SUMMARY.md) with relevant findings for score ≥ 7; include notes for not-relevant papers (score < 7).
- Ensure DOIs and PMIDs are presented as clickable links in SUMMARY.md.
- If medicinal chemistry is involved, use
When to use it
After literature search returns results. When need to determine if paper contains specific data. When screening papers for relevance. When extracting methods, results, data from papers.
What it can touch
Tools include claude-code for code-related steps and checks, and scripts for ChEMBL checks, full-text fetching, and data extraction as described.
Caveats
License: NOASSERTION. The skill relies on paywall handling via Unpaywall; if no OA version is found, abstract-only evaluation proceeds. It requires reporting for every paper and mandates adding all papers to papers-reviewed.json and updating SUMMARY.md accordingly.
<!-- ╔══════════════════════════════════════════════════════════════╗ ║ 本文件为开源 Skill 原始文档,收录仅供学习与研究参考 ║ ║ CoPaper.AI 收集整理 | https://copaper.ai ║ ╚══════════════════════════════════════════════════════════════╝ 来源仓库: https://github.com/kthorn/research-superpower 项目名称: research-superpower 开源协议: MIT License 收录日期: 2026-04-02 声明: 本文件版权归原作者所有。此处收录旨在为社会科学实证研究者 提供 AI Agent Skills 的集中参考。如有侵权,请联系删除。 --> # E
What does the Evaluating Paper Relevance skill do?
Two-stage paper screening - abstract scoring then deep dive for specific data extraction
How do I install it?
Run `npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill evaluating-paper-relevance --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Auto-Empirical-Research-Skills, a repository with 3,244 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.