# Arxiv

> arXiv

- Skill: `spike-faye-lei/arxiv` (Agent Skill, multi-file: 9 files)
- Install (CLI): `npx skillmds@latest add spike-faye-lei/arxiv`
- Raw SKILL.md: https://api.skillmd.com/api/skills/spike-faye-lei/arxiv/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Research & Search
- Author: spike-faye-lei (https://skillmd.com/u/spike-faye-lei)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/spike-faye-lei/arxiv

---

# arXiv

## Overview
Open-access preprint server for scientific papers (arxiv.org). Public Atom API at export.arxiv.org for search and metadata retrieval, plus HTML abstract pages.

## Workflows

### Find papers on a topic
1. `searchPapers(search_query)` → `id`, `title`, `summary`, `category` per entry
2. `getPaper(id_list)` → full metadata with `title`, `authors`, `abstract`, `links` (PDF link)

### Get a specific paper's details
1. `getPaper(id_list: "1706.03762")` → `title`, `authors`, `abstract`, `categories`, `links`
2. Extract PDF link from `<link title="pdf">` → `href`

### Quick abstract lookup
1. `getAbstract(arxiv_id: "1706.03762")` → `title`, `authors`, `abstract`

## Operations

| Operation | Intent | Key Input | Key Output | Notes |
|-----------|--------|-----------|------------|-------|
| searchPapers | find papers by keyword/author/category | search_query (user) | XML feed with entries (id, title, authors, abstract, categories) | **entry point** — paginated via start/max_results, field prefixes (all:, au:, ti:, cat:) |
| getPaper | get paper metadata by ID | id_list ← searchPapers `<id>` or user | XML entry with full metadata and PDF link | can fetch multiple papers at once |
| getAbstract | get abstract page by ID | arxiv_id ← searchPapers `<id>` or user | HTML page with title, authors, abstract | human-readable format, uses arxiv.org host |

## Quick Start

```bash
# Search for papers about transformers
openweb arxiv exec searchPapers '{"search_query": "all:transformer", "max_results": 5}'

# Get metadata for "Attention Is All You Need"
openweb arxiv exec getPaper '{"id_list": "1706.03762"}'

# Get abstract page for a paper
openweb arxiv exec getAbstract '{"arxiv_id": "1706.03762"}'
```