skills.rest
huangrt01
huangrt01COMMUNITY
Explore all skills by this author →
GitHub

research-material-scout

Researches, triages, and archives learning materials from papers, repos, and social sources.

4.0k|479|Updated Apr 16, 2020
Search Tags:#note taking#literature search#source verification#Research#paper reading#Arxiv#material intake
One-click install
npx skills add https://github.com/huangrt01/CS-Notes --skill research-material-scout
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: research-material-scout
Source: https://github.com/huangrt01/CS-Notes/tree/main/.codex/skills/research-material-scout
Command: npx skills add https://github.com/huangrt01/CS-Notes --skill research-material-scout

SYSTEM DOCUMENTATION & REQUIREMENTS

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

What problem does it solve?

It turns scattered links, papers, and social posts into a verified, ranked learning-material pipeline, so research intake, paper reading, and note integration follow a consistent evidence-based workflow instead of ad-hoc bookmarking.

Core Features & Use Cases

  • Material intake and triage: Processes 素材： links by reading the source, classifying it into S/A/B/Unread tiers, and recording candidates with exact-read evidence and receipts.
  • Multi-source paper exploration: Runs parallel recall across arXiv, OpenReview, OpenAlex, bioRxiv, medRxiv, ChemRxiv, and venue search lanes via a standard-library Python script, then merges results with web and repo evidence.
  • Structured paper reading: Applies a Keshav three-pass protocol for 请你读 / 精读 requests, producing claim maps, mechanism summaries, artifact deltas, and reader maps.
  • Use Case: A user sends 调研：agent memory benchmarks; the skill drafts intent profiles, queries multiple paper lanes plus GitHub and docs, ranks candidates with evidence sentences, and outputs a deep/quick/background/carryover reading split.

Quick Start

Ask the assistant to research a topic or read a link, for example by saying "调研：latest agent memory papers" or "请你读 this arXiv paper", and it will gather sources, verify them, and return a ranked summary.

Frequently Asked Questions about research-material-scout

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

FAQPage Schema
How do I search multiple paper sources like arXiv and OpenReview at once?▼

Run the multi_source_paper_explore.py script with repeated --query flags and a --sources list such as arxiv,openreview,openalex. It queries each lane in parallel, normalizes metadata into JSON, and deduplicates results across lanes for downstream ranking.

How to read an arXiv paper thoroughly with an AI assistant?▼

Use the 请你读 or 精读 directive, which triggers a three-pass protocol: triage via abstract and figures, targeted content reading, and selective virtual reimplementation. The skill prefers arXiv HTML, then TeX source, with PDF only as fallback.

Does the paper exploration script need external Python packages?▼

No, the script uses only Python standard library modules such as urllib, json, and xml parsing. It calls public APIs from arXiv, OpenReview, OpenAlex, bioRxiv, medRxiv, and a ChemRxiv data mirror directly.

What are the limitations of social media as a research source?▼

Social platforms like X, 小红书, and 微信公众号 are treated as discovery signals, not final truth. Technical claims must be traced back to primary sources such as papers, repositories, or official docs before receiving an S or A ranking.

When should I use 整理笔记 instead of 素材 intake?▼

Use 整理笔记 when you want content integrated directly into a named note in the Notes directory rather than added to the candidate library. The named target and the source's primary domain take priority over career-based routing.

Related Skills

Explore More Education & Research Skills→
laik184laik184

podcast-marketing

Extract content atoms from podcast transcripts and generate marketing assets.

Community
Intermediate
KeshariPiyush24KeshariPiyush24

Interview Prep Generator

Generate STAR stories and interview questions from resume and role context.

Community
Intermediate
confighubconfighub
14

pilot-rollback-decision

Evaluate ConfigHub history and receipts to select the safest Kubernetes rollback revision.

Official
Advanced
AradotsoAradotso
7

modular-rag-mcp-server

Deploy a modular RAG pipeline with hybrid retrieval via MCP.

Official
Advanced