What is the Web Fetch skill for Claude Code?

It is a specialized command that allows Claude to use the Gemini CLI to fetch web content, convert it to clean Markdown, and use it as context for coding tasks.

Does this skill require external dependencies?

Yes, you must have the Gemini CLI installed globally via npm (npm i -g @google/gemini-cli) for this skill to function.

Why should I use this instead of native WebFetch?

This skill serves as a battle-tested alternative for instances where Claude's native tool fails, or when you need more control over what specific parts of a page are extracted.

Can I filter the content extracted from a URL?

Yes, the skill supports optional extraction instructions, allowing you to tell the AI to focus only on specific elements like API tables or code snippets.

Web Fetch via Gemini CLI

Name: Web Fetch via Gemini CLI
Author: Dokkabei97

byDokkabei97

•

ウェブスクレイピングとデータ収集

Extracts web page content into clean Markdown using the Gemini CLI as a robust alternative to native browsing tools.

The Web Fetch skill provides a resilient, enterprise-grade fallback for Claude Code's native web browsing capabilities. By leveraging the Gemini CLI and specialized prompt engineering, it retrieves web content while stripping away noise like advertisements, navigation bars, and footers. This skill is particularly effective for developers who need to ingest technical documentation, API references, or GitHub READMEs directly into their workflow, providing structured Markdown that preserves code blocks and tables even when standard fetch tools encounter access restrictions.

主な機能

01High-speed performance using the gemini-3.1-flash-preview model

02Clean Markdown output preserving tables, lists, and code blocks

03Custom extraction instructions for targeted content retrieval

04Reliable fallback mechanism for failed native WebFetch attempts

052 GitHub stars

06Intelligent noise reduction removing ads, footers, and sidebars

ユースケース

01Retrieving specific API documentation tables directly into the terminal

02Scraping GitHub repository files and READMEs for context

03Ingesting technical blog posts for code implementation reference

主な機能

01High-speed performance using the gemini-3.1-flash-preview model

02Clean Markdown output preserving tables, lists, and code blocks

03Custom extraction instructions for targeted content retrieval

04Reliable fallback mechanism for failed native WebFetch attempts

052 GitHub stars

06Intelligent noise reduction removing ads, footers, and sidebars

ユースケース

01Retrieving specific API documentation tables directly into the terminal

02Scraping GitHub repository files and READMEs for context

03Ingesting technical blog posts for code implementation reference