arxiv-paper-get

arxiv-paper-get

Download an arXiv paper — PDF + LaTeX source — in one command.

Given an arXiv URL, arxiv-paper-get fetches the paper's PDF and LaTeX source tarball in parallel, extracts the source into a source/ directory, and creates a ready-to-use local workspace with metadata and a Markdown report skeleton.

GitHub: hao1305430263/arxiv-paper-get

---

Features

• Parallel download — PDF and LaTeX source (.tar.gz) are fetched simultaneously.
• Auto-extract — the source tarball is unpacked into source/ with path-traversal protection.
• Rich metadata — title, authors, abstract, categories, DOI, journal ref, and more are saved as metadata.json (sourced from the official arXiv API).
• Report skeleton — a structured Markdown report is created if one doesn't already exist (never overwrites).
• Zero dependencies — uses only the Python standard library. No pip install of anything else required.
• Cross-platform — pathlib.Path throughout; Windows and Unix paths are handled correctly.
• Graceful degradation — if LaTeX source is not available (common for older or some publisher-hosted papers), the PDF is still downloaded and the tool exits cleanly.

---

Installation

Option A: Claude Code plugin marketplace (recommended)

In Claude Code, add this repo as a plugin marketplace, then install:

/plugin marketplace add hao1305430263/arxiv-paper-get
/plugin install arxiv-paper-get@arxiv-paper-get

This installs the Claude Code plugin, which carries the arxiv-paper-get skill. Once installed, the skill auto-triggers whenever you share an arXiv link.

Option B: npx skills (cross-agent)

npx skills add https://github.com/hao1305430263/arxiv-paper-get

3. Install the Python CLI

Regardless of which option you used above, also install the CLI tool:

uv tool install git+https://github.com/hao1305430263/arxiv-paper-get

Uninstall

uv tool uninstall arxiv-paper-get

---

Usage

# Default: saves to ./papers/
arxiv-paper-get "https://arxiv.org/abs/2506.01966"

# Custom base directory
arxiv-paper-get "https://arxiv.org/abs/2506.01966" "/path/to/my/papers"

Output

The command prints a JSON summary to stdout:

{
  "arxiv_id": "2506.01966",
  "title": "Matrix Is All You Need",
  "paper_dir": "papers/Matrix_Is_All_You_Need",
  "pdf_path": "papers/Matrix_Is_All_You_Need/Matrix_Is_All_You_Need.pdf",
  "source_path": "papers/Matrix_Is_All_You_Need/Matrix_Is_All_You_Need.tar.gz",
  "source_extract_dir": "papers/Matrix_Is_All_You_Need/source",
  "report_path": "papers/Matrix_Is_All_You_Need/Matrix_Is_All_You_Need_report.md",
  "metadata_path": "papers/Matrix_Is_All_You_Need/metadata.json",
  "pdf_ok": true,
  "source_ok": true,
  "source_extracted": true
}

Directory structure created

papers/{Paper Title}/
├── {Paper Title}.pdf          # PDF file
├── source/                    # Extracted LaTeX source (if available)
│   ├── main.tex
│   ├── sections/
│   └── ...
├── metadata.json              # Full paper metadata
└── {Paper Title}_report.md    # Report skeleton for note-taking

The LaTeX source tarball (.tar.gz) is auto-cleaned after successful extraction.

---

Credits

Based on the paper-interpreter skill by chujianyun.

---

---

arxiv-paper-get（中文说明）

一条命令下载 arXiv 论文 —— PDF + LaTeX 源码。

给定一个 arXiv 链接，arxiv-paper-get 会并行下载论文的 PDF 和 LaTeX 源码压缩包， 自动将源码解压到 source/ 目录，并创建包含元数据和 Markdown 报告骨架的本地工作区。

GitHub: hao1305430263/arxiv-paper-get

---

功能特性

• 并行下载 — PDF 和 LaTeX 源码（.tar.gz）同时获取，互不等待。
• 自动解压 — 源码包自动解压到 source/ 子目录，内置路径穿越保护。
• 完整元数据 — 标题、作者、摘要、分类、DOI、期刊引用等信息保存为 metadata.json（通过 arXiv 官方 API 获取）。
• 报告骨架 — 自动生成结构化 Markdown 报告模板，已有报告不会被覆盖。
• 零依赖 — 仅使用 Python 标准库，无需额外安装任何第三方包。
• 跨平台 — 全面使用 pathlib.Path，Windows 与 Unix 路径均正确处理。
• 优雅降级 — 如果论文的 LaTeX 源码不可用（老旧论文或部分出版社论文常见）， 仍然会下载 PDF 并正常退出。

---

安装

方式 A：Claude Code 插件市场（推荐）

在 Claude Code 中，先将本仓库添加为插件市场，然后安装：

/plugin marketplace add hao1305430263/arxiv-paper-get
/plugin install arxiv-paper-get@arxiv-paper-get

这会安装一个 Claude Code 插件，其中包含 arxiv-paper-get skill。 安装后，当你向 Claude Code 分享 arXiv 链接时，skill 会自动触发。

方式 B：npx skills（跨 Agent）

npx skills add https://github.com/hao1305430263/arxiv-paper-get

3. 安装 Python CLI

无论选择哪种方式，都需要额外安装命令行工具：

uv tool install git+https://github.com/hao1305430263/arxiv-paper-get

卸载

uv tool uninstall arxiv-paper-get

---

用法

# 默认保存到 ./papers/
arxiv-paper-get "https://arxiv.org/abs/2506.01966"

# 指定保存目录
arxiv-paper-get "https://arxiv.org/abs/2506.01966" "/path/to/my/papers"

输出

命令执行后在 stdout 输出 JSON 摘要（见上方英文部分）。

生成的目录结构