Skip to content
Back to skills

Social Media Paper Triage

ASecurity

Extracts paper recommendations from social-media posts and online articles (小红书 / Xiaohongshu / RedNote, 微信公众号 / WeChat Official Accounts, Twitter / X threads, Reddit posts, Bilibili videos, blog posts, newsletters, Jina Reader URLs), identifies the underlying academic papers, locates the authoritative original sources (arXiv, conference proceedings, DOI), and triages relevance to the user's research before any library action. Use when the user forwards a social-media link, screenshot, or art...

  • 67 stars
  • 0 votes
  • 0 copies
  • 36 views
  • Added May 27, 2026
researchpythongo

Works with

  • mcp

Security analysis

A100/100

Scanned May 27, 2026

npx -y skills add jxtse/scientific-research-skills --skill social-media-paper-triage --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Social Media Paper Triage?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Social Media Paper Triage
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/jxtse-social-media-paper-triage/badge)](https://www.skillsdirectory.com/skills/jxtse-social-media-paper-triage)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: social-media-paper-triage
description: >
  Extracts paper recommendations from social-media posts and online articles
  (小红书 / Xiaohongshu / RedNote, 微信公众号 / WeChat Official Accounts,
  Twitter / X threads, Reddit posts, Bilibili videos, blog posts, newsletters,
  Jina Reader URLs), identifies the underlying academic papers, locates the
  authoritative original sources (arXiv, conference proceedings, DOI), and
  triages relevance to the user's research before any library action. Use when
  the user forwards a social-media link, screenshot, or article that mentions
  a paper / method / model, asks to "find the original paper" from a blog or
  thread, shares a 调研贴 / 论文推荐 / paper recommendation post, or wants
  to evaluate whether a buzz-worthy paper is worth reading before adding it
  to Zotero.
---

# Social Media Paper Triage

Turn social media paper recommendations into actionable research items.

## When to Use

- User forwards a 小红书 post about a paper
- User shares a WeChat公众号 article discussing papers
- User shares a Twitter/X thread about a paper or method
- User asks "find the original paper from this link"
- User shares any blog post / newsletter that references academic papers

## Workflow

### Step 1: Extract Content from Platform

Use platform-specific tools to fetch the full content:

| Platform | Tool | Command |
|----------|------|---------|
| **小红书** | Agent Reach (XiaoHongShu) | `mcporter call 'xhs.get_note(note_url: "URL")'` |
| **WeChat公众号** | Agent Reach (WeChat) | `python3 ~/.agent-reach/.venv/bin/wechat_article.py "URL"` |
| **Twitter/X** | xreach | `xreach tweet URL --json` or `xreach thread URL --json` |
| **Reddit** | Agent Reach | `mcporter call 'reddit.read_post(url: "URL")'` |
| **Bilibili** | Agent Reach | `mcporter call 'bilibili.get_video_info(url: "URL")'` |
| **Any URL** | Jina Reader | `curl -s "https://r.jina.ai/URL"` |

### Step 2: Identify Papers

From the extracted content, identify all referenced papers:
- Look for: paper titles, arXiv IDs, DOIs, author names + year citations
- Distinguish between: the main paper being discussed vs. papers cited in passing
- Note: social media posts often use informal titles or translated titles

### Step 3: Find Original Sources

For each identified paper, find the authoritative source:

1. **arXiv search**: Check if it's on arXiv (most ML/AI papers are)
2. **Semantic Scholar**: Search by title for metadata + citation count
3. **Google Scholar** (via web search): Fallback for non-arXiv papers

Priority: arXiv PDF > conference proceedings > journal version

### Step 4: Summarize for Decision

Present a concise summary to the user:

```
📄 Paper: [Title]
👥 Authors: [First author] et al., [Year]
🏛 Venue: [Conference/Journal]
📊 Citations: [N]
🔗 Original: [arXiv/DOI link]
📱 Source: [social media link]

TL;DR: [2-3 sentence summary of what the paper does and why it matters]

Relevance to your work: [brief assessment based on user's research context]
```

### Step 5: User Decision → Action

Based on user's response:
- **"Add to reading queue"** → Add to Zotero `30_Reading Queue` with appropriate tag
- **"Not relevant"** → Done, no action
- **"Read it now"** → Switch to paper-reading skill
- **"Save for later"** → Add to Zotero with lower priority tag

## Key Principles

1. **Always find the original paper** — don't just summarize the social media post
2. **Don't auto-add to Zotero** — summarize first, let user decide
3. **Preserve the social media link** — add as a note/attachment in Zotero for provenance
4. **Assess relevance** — use knowledge of user's active projects to judge fit
5. **Batch processing** — if the post mentions multiple papers, triage all of them at once

## Common Gotchas

- WeChat articles often translate paper titles to Chinese — search by original English title
- 小红书 posts may summarize methods inaccurately — always verify against the original
- Twitter threads may reference preprints that have been updated/published since
- Some posts discuss methods without naming specific papers — ask user for clarification

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.