Agent Skills › Prismer-AI/Prismer › paper-search

paper-search

GitHub

用于搜索和发现 arXiv、Semantic Scholar 等学术数据库中的论文，支持按主题、作者或引用检索，提供元数据整理及 PDF 阅读功能。

docker/templates/academic-researcher/skills/paper-search/SKILL.md Prismer-AI/Prismer

Trigger Scenarios

查找研究论文搜索 arXiv 文献综述获取学术参考文献

Install

npx skills add Prismer-AI/Prismer --skill paper-search -g -y

More Options

Non-standard path

npx skills add https://github.com/Prismer-AI/Prismer/tree/main/docker/templates/academic-researcher/skills/paper-search -g -y

Use without installing

npx skills use Prismer-AI/Prismer@paper-search

指定 Agent (Claude Code)

npx skills add Prismer-AI/Prismer --skill paper-search -a claude-code -g -y

安装 repo 全部 skill

npx skills add Prismer-AI/Prismer --all -g -y

预览 repo 内 skill

npx skills add Prismer-AI/Prismer --list

SKILL.md

Frontmatter

{
    "name": "paper-search",
    "description": "Search and discover academic papers from arXiv and web sources using arxiv_to_prompt and web search. Use when the user asks to find research papers, search arXiv, look up scientific literature, do a literature review, or needs academic references."
}

Paper Search Skill

Description

Search and discover academic papers from various databases and repositories.

Tools Used

arxiv_to_prompt - Convert arXiv papers to LLM-readable text (search by arXiv ID)
load_pdf - Load PDF papers in the workspace viewer
Web search is available as a built-in agent capability

Capabilities

Search Academic Databases

arXiv (preprints in physics, math, CS, etc.)
Semantic Scholar (comprehensive academic search)
Google Scholar (broad academic coverage)
PubMed (biomedical literature)

Retrieve Paper Information

Title, authors, abstract
Publication venue and date
Citation count and references
Download links (when available)

Organize Results

Filter by date, citation count, relevance
Group by topic or author
Export to reference managers

Usage Patterns

Basic Search

When user says: "Find papers about transformer architectures"

Formulate search query
Search primary databases (arXiv, Semantic Scholar)
Collect top results with metadata
Present organized list with key information

Specific Author Search

When user says: "Find recent papers by [Author Name]"

Search by author name
Filter by publication date
Show papers with co-authors and venues

Citation Search

When user says: "Find papers that cite [Paper Title]"

Identify the source paper
Search for citing papers
Organize by recency and relevance

Related Work Search

When user says: "Find related work for this topic"

Identify key concepts and terms
Search across multiple databases
Categorize by methodology or approach
Highlight highly cited foundational papers

Output Format

Single Paper Result

Title: [Paper Title]
Authors: [Author List]
Year: [Publication Year]
Venue: [Journal/Conference]
Abstract: [First 200 words...]
Citations: [Count]
Link: [URL]

Search Results Summary

Found [N] relevant papers:

Top Results:
1. [Title] ([Year]) - [Citation Count] citations
2. [Title] ([Year]) - [Citation Count] citations
...

Would you like me to:
- Show detailed abstracts?
- Filter by specific criteria?
- Export to BibTeX?

Tool Examples

Read an arXiv paper

arxiv_to_prompt arxiv_id="2303.08774"

Get just the abstract

arxiv_to_prompt arxiv_id="2303.08774" abstract_only=true

List paper sections before full read

arxiv_to_prompt arxiv_id="2303.08774" list_sections=true

Load a PDF paper

load_pdf source="/workspace/papers/downloaded.pdf"

Best Practices

Start with abstracts: Use abstract_only=true to quickly triage papers before committing to a full read
Multiple Sources: Cross-reference across databases for comprehensive coverage
Citation Awareness: Note highly-cited foundational works
Verify before reading: Use list_sections=true to check a paper's structure before fetching full text