gemini-research

Orchestrate Gemini CLI queries with Google search grounding to generate cited JSON and Markdown research reports.

4|Updated Oct 6, 2025
One-click install
npx skills add https://github.com/otrebu/agents --skill gemini-research
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gemini-research
Source: https://github.com/otrebu/agents/tree/main/plugins/knowledge-work/skills/gemini-research
Command: npx skills add https://github.com/otrebu/agents --skill gemini-research

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Performs real-time, grounded web research via Gemini CLI and produces structured markdown reports for Claude to analyze and summarize.

Core Features & Use Cases

  • Modes: quick, deep, and code-focused research for varying depth.
  • Structured Output: JSON and Markdown with sources, key points, quotes, and summaries.
  • Critical Workflow: Claude must append its analysis to the generated report.

Quick Start

bash plugins/knowledge-work/skills/gemini-research/scripts/research.sh "your query" [quick|deep|code]

Frequently Asked Questions about gemini-research

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I perform web research with citations using the command line?

Web research with citations is performed by orchestrating Gemini CLI queries grounded in Google search results. This Skill automates that workflow via a bash script that executes your query in one of three modes—quick, deep, or code—and outputs structured JSON and Markdown reports with traceable sources, key points, and quotes.

Can I use Gemini CLI to generate structured research reports with sources?

Yes. Gemini CLI grounds queries in real-time Google search results and this Skill wraps that capability to produce JSON and Markdown reports. Each report includes citations, key findings, and quotes, enabling you to trace claims back to their sources.

What's the difference between quick, deep, and code research modes?

The three modes adjust research depth and focus: quick mode performs rapid topic overview, deep mode investigates thoroughly across multiple angles, and code mode targets code patterns and technical implementation examples. All modes produce cited, structured output.

Do I need Claude to analyze the research output?

Yes. The Skill's workflow mandates post-analysis augmentation by Claude after the Gemini CLI generates the grounded report. Claude appends its analysis and summary to the Markdown output, completing the research-to-insight pipeline.

What topics can I research with this Skill?

Web research covers news, documentation, tutorials, and code patterns across any topic. The Skill applies to both fast research workflows and deep investigations. Gemini CLI grounds queries in current Google search results, so topical scope is unrestricted by the Skill itself.

What output formats does the research produce?

Research output is structured as JSON with citations, key points, and quotes, plus a Markdown report. Both formats preserve source traceability and are designed for Claude to read, analyze, and augment with additional insights before final delivery.