arxiv-search

arXiv Search

Search arXiv for papers by query, author, or category, with option to process results.

Arguments

Parse from: $ARGUMENTS

• QUERY: Search query (required)

Options:

• --author "NAME": Filter by author name
• --category CAT: Filter by arXiv category (e.g., cs.LG, cs.CL, cs.AI)
• --max N: Maximum results to return (default: 10)
• --sort relevance|date: Sort order (default: relevance)
• --process N: Automatically process top N papers with NotebookLM
• --days N: Only papers from last N days

Categories Reference

Common ML/AI categories:

• cs.LG - Machine Learning
• cs.CL - Computation and Language (NLP)
• cs.AI - Artificial Intelligence
• cs.CV - Computer Vision
• cs.NE - Neural and Evolutionary Computing
• stat.ML - Machine Learning (Statistics)

Workflow

1. Construct arXiv search URL with parameters
2. Navigate to arXiv or use API
3. Extract search results:
   • Title
   • Authors
   • arXiv ID
   • Abstract
   • Submission date
   • Categories
4. Display formatted results
5. If --process specified, create NotebookLM notebooks for top papers

Example Usage

/notebooklm-papers:arxiv-search "transformer architecture"
/notebooklm-papers:arxiv-search "large language models" --category cs.CL --max 20
/notebooklm-papers:arxiv-search "reinforcement learning" --days 7 --process 3
/notebooklm-papers:arxiv-search --author "Yann LeCun" --max 5

Output Format

# arXiv Search Results

Query: "{query}"
Category: {category}
Results: {count}

## Papers

### 1. {Title}
- **arXiv**: [{id}](https://arxiv.org/abs/{id})
- **Authors**: {authors}
- **Date**: {date}
- **Categories**: {categories}

> {abstract preview...}

[PDF](https://arxiv.org/pdf/{id}.pdf) | [Process with NotebookLM](#)

---
...