Skill: arxiv-to-md

Version: 1.0.0

Status: Active

required_canon_version: ">=1.0.0"

Trigger

When the user asks to download, convert, or fetch an arXiv paper as markdown. Examples:

• "Download arxiv 1706.03762"
• "Get this paper as markdown: https://arxiv.org/abs/2401.12345"
• "Convert arxiv paper 1510.00149 to md"

Inputs

Outputs

• Markdown file with proper #, ##, ### heading structure
• Prints confirmation with output path

Methods

Execution

# Run via venv python
.venv/Scripts/python.exe CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py <arxiv_id> <output_path> [method]

Examples

# Auto method (recommended)
.venv/Scripts/python.exe CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py 1706.03762 output.md

# Force HTML (no pandoc needed)
.venv/Scripts/python.exe CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py 1706.03762 output.md html

# Force LaTeX (best quality)
.venv/Scripts/python.exe CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py 1706.03762 output.md latex

# From URL
.venv/Scripts/python.exe CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py https://arxiv.org/abs/1706.03762 output.md

Constraints

• Network access required to fetch from arxiv.org / ar5iv.labs.arxiv.org
• LaTeX method requires pandoc (C:\Users\<user>\AppData\Local\Pandoc\pandoc.exe)
• Some papers have non-standard LaTeX that pandoc can't parse (auto falls back to HTML)
• Output is UTF-8 encoded markdown

Dependencies

Python packages (in .venv):

• requests
• markdownify
• beautifulsoup4

System (for latex method):

• pandoc - installed via winget install JohnMacFarlane.Pandoc

Error Handling

Implementation

Script: CAPABILITY/SKILLS/utilities/arxiv-to-md/pdf_converter.py

Key functions:

• convert_arxiv(arxiv_input, output_path, method) - Main entry point
• convert_arxiv_latex(arxiv_id) - LaTeX + pandoc method
• convert_arxiv_html(arxiv_id) - ar5iv HTML method
• parse_arxiv_id(arxiv_input) - Extract ID from URL or raw input