Where does the skill store the downloaded papers?

Original source files are stored in a global cache directory, while the generated summaries are saved directly into your project's local knowledge folder for easy reference.

Does it re-download the paper every time I ask?

No, the skill includes a caching mechanism that checks for existing local copies of the ArXiv ID before attempting a new download.

Can it handle papers split across multiple .tex files?

Yes, the skill is designed to identify the main entry point (like main.tex) and recursively read through all referenced source files.

Why does this skill download TeX source instead of a PDF?

LaTeX source provides structured text and raw mathematical notation that is much easier for AI to parse accurately compared to the visual layout of a PDF.

ArXiv Paper Reader

Name: ArXiv Paper Reader
Author: chrisjmccormick

bychrisjmccormick

•

Data Science & ML

Downloads, parses, and summarizes ArXiv research papers by analyzing their raw LaTeX source code for deep technical insights.

This skill empowers Claude to ingest scientific research from ArXiv by fetching raw LaTeX source files instead of standard PDFs, ensuring significantly higher accuracy when reading complex equations and document structures. It automates the entire workflow of normalizing ArXiv URLs, downloading and unpacking source archives, and recursively traversing multi-file TeX projects. Designed for researchers and developers, it generates project-aware summaries that bridge the gap between academic theory and practical implementation, specifically focusing on how research findings can be applied to your current codebase.

Key Features

01Smart URL normalization to ensure the correct source files are retrieved

02Fetches raw LaTeX source for superior parsing accuracy over PDF content

03Automatically handles .tar.gz extraction and local knowledge caching

04Recursively traverses multi-file TeX projects to capture full context

051 GitHub stars

06Generates context-aware markdown summaries with project-specific takeaways

Use Cases

01Extracting optimization techniques from academic papers for performance tuning

02Understanding new machine learning architectures to implement them in a codebase

03Building a local knowledge base of research relevant to a specific software project

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add chrisjmccormick/modded-nanochat read-arxiv-paper

For use in Claude.ai and ChatGPT

Download Skill