How long does a research task typically take?

Most deep research tasks are completed within 5 to 15 minutes, depending on the complexity of the topic and the depth of analysis requested.

How does prompt optimization work?

The skill takes your initial query and applies a template that defines scope, output structure, and source requirements to ensure Gemini provides the most comprehensive and verifiable data possible.

Do I need a separate API key for this skill?

Yes, this skill requires a GEMINI_API_KEY environment variable configured in your shell to access the Gemini Interactions API.

Can I use this for searching my local codebase?

No, this skill is specifically designed for real-time web research. For local codebase questions, you should use Claude Code's native file analysis capabilities.

Where are the research reports saved?

Results are automatically saved as Markdown files in your project directory, typically under a references folder, for easy documentation and review.

Gemini Deep Research

Name: Gemini Deep Research
Author: cdeistopened

bycdeistopened

•

웹 스크래핑 및 데이터 수집

Executes comprehensive, multi-source web research and analysis using Gemini's Deep Research agent directly from your terminal.

This skill integrates Gemini's powerful Deep Research capabilities into Claude Code, enabling users to perform exhaustive information gathering across 50-100+ web sources. It features an automated prompt optimization engine that transforms simple queries into structured research plans, manages asynchronous background execution via the Gemini Interactions API, and delivers formatted Markdown reports with verified citations. Whether you need competitive landscaping, technical literature reviews, or market due diligence, this skill handles the heavy lifting of data collection and synthesis while you continue your development workflow.

주요 기능

01Built-in anti-hallucination guardrails and source verification

02Automated prompt optimization to maximize research depth and structure

03Asynchronous background execution with real-time status polling

04Comprehensive multi-source analysis across 50-100+ web citations

05Structured report generation with executive summaries and comparison tables

068 GitHub stars

사용 사례

01Performing technical literature reviews and state-of-the-art surveys

02Conducting market research and competitive landscaping for new software features

03Executing deep-dive due diligence on emerging technologies or industry movements

주요 기능

01Built-in anti-hallucination guardrails and source verification

02Automated prompt optimization to maximize research depth and structure

03Asynchronous background execution with real-time status polling

04Comprehensive multi-source analysis across 50-100+ web citations

05Structured report generation with executive summaries and comparison tables

068 GitHub stars

사용 사례

01Performing technical literature reviews and state-of-the-art surveys

02Conducting market research and competitive landscaping for new software features

03Executing deep-dive due diligence on emerging technologies or industry movements