Skip to content
Get packs

arXiv Research Assistant

Search, read, cite and download arXiv papers and pull citation data from Semantic Scholar.

Source Download skill
Repository
XiaomiMiMo/MiMo-Code
Stars
13,743
Licence
MIT
Files
2 files

What it does

Once installed, the agent can search the arXiv preprint archive by topic, author or category, pull abstracts and metadata, generate BibTeX citations, download PDFs, list the newest papers in a field, and look up who cites a paper, what it references, and related work via Semantic Scholar. It uses a bundled Python script for these lookups, so results come back in a clean, readable or JSON form rather than raw XML.

Great for

  • Students and researchers doing a literature review
  • People who want citations in BibTeX format without hunting for them
  • Anyone tracking the latest submissions in a field such as cs.AI or cs.CL

Skip it if

  • Finding paywalled journal articles, since it only covers arXiv and Semantic Scholar open metadata; citation counts may be incomplete or lag behind
  • Verifying that a paper's claims are correct, since the skill retrieves papers and summaries but does not check their quality or reproducibility
  • Tasks needing private data, as it sends search terms and paper IDs to public external services

Try asking

  • “Find the most recent papers on GRPO reinforcement learning and summarize their abstracts”
  • “Generate BibTeX entries for arXiv 1706.03762 and 2601.02780”
  • “Show who cites this paper and which related papers I might have missed”
  • “List the latest 10 submissions in cs.CL and download the PDFs I select”
Step-by-step guide: how to use arXiv Research AssistantInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    arxiv.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip arxiv.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip arxiv.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Ask for specific arXiv IDs when you can, since they give exact results and preserve version numbers for citations
  • Expect rate limits (about one request every 3 seconds for arXiv and one per second for Semantic Scholar) and allow the agent to space out large batches
  • Check any paper flagged [WITHDRAWN] before citing it, and verify key claims by reading the full text yourself

What's inside

  • scripts/arxiv.py
  • SKILL.md

Redistributed unmodified from XiaomiMiMo/MiMo-Code under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.