Skip to content

Gemini Literature Search

Find research papers broadly via Google Gemini, beyond arXiv and Semantic Scholar keyword search

Stars
18,482
Licence
MIT
Files
1 file

What it does

When you ask for papers on a topic, the agent sends a detailed search request to Google's Gemini AI through a bridge tool. Gemini breaks the topic into sub-problems and alternative names, then returns a numbered list of candidate papers with authors, year, venue, arXiv ID, DOI and code links. It is meant to surface papers that plain keyword searches miss, but Gemini's answers are generated by a language model and are not checked against a database, so every citation needs verifying.

Great for

  • Researchers starting a literature review who want broad coverage across topic variants
  • Students mapping a new field and looking for surveys and related work
  • Anyone who already has Gemini CLI set up and wants a second discovery source alongside arXiv or Semantic Scholar

Skip it if

  • Final bibliographies or citations without checking: Gemini can invent or misattribute titles, authors, years or DOIs, so each paper must be verified at the source
  • Users without a Google account or Gemini API key, since the skill cannot run without that setup

Try asking

  • “gemini search diffusion models for medical image segmentation”
  • “search with gemini contrastive learning for speech, year: 2021-”
  • “gemini papers on graph neural networks for drug discovery, code-only, max: 20”
Step-by-step guide: how to use Gemini Literature SearchInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    gemini-search.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

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

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip gemini-search.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

  • Verify every returned paper on arXiv, the publisher site or Semantic Scholar before citing it, because the output is model-generated text.
  • Set GEMINI_API_KEY in the shell before launching the agent, since running processes do not pick up later exports or .env files automatically.
  • Use the model override only if you understand the quota and downgrade behaviour described in the file; the default auto-gemini-3 setting routes between Gemini 3 models by server load.

What's inside

  • SKILL.md

Redistributed unmodified from wanshuiyin/Auto-claude-code-research-in-sleep 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.