AI Agent Hub
Back to skills
Web Fetch Toolkit icon

Web Fetch Toolkit

Data Analysis Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please follow https://skillhub.cn/install/skillhub.md to install @user_f28fdad8/oc-web-fetch.

About this skill

Problem

When feeding web content to LLMs or analysis pipelines, raw scraping quickly requires handling encoding, main-content noise, batch deduplication, element extraction, and Markdown conversion. @user_f28fdad8/oc-web-fetch reduces that workflow to four scripts: single-URL fetch, batch fetch, selector extraction, and HTML-to-Markdown conversion.

How It Works

fetch.py issues an HTTP request, detects encoding, extracts the main text, and returns title, body, and metadata. batch_fetch.py fetches multiple URLs through a concurrent queue and reports success/failure counts plus elapsed time. extract.py supports tag, class, id, and attr selectors, returning only matched text or attributes. to_markdown.py removes navigation, ads, scripts, and styles, then preserves headings, paragraphs, lists, links, and images as clean Markdown. The inputs and outputs are standardized as JSON where possible, making it easier for agents or downstream programs to consume.

Boundaries

It is best for static pages and ordinary HTML body extraction. Its text-density-based cleaning does not replace browser rendering, login sessions, JavaScript-heavy UIs, or complex interaction flows. Batch fetching uses conservative default concurrency and delay; in production, check target-site robots, rate limits, and compliance constraints.

Use Cases

  • Fetch one competitor page, extract title, body, and metadata, then convert it to Markdown for model analysis
  • Fetch multiple source URLs concurrently and produce a summary of success/failure counts plus elapsed time
  • Use class or id selectors to extract a target section and store the matched text or attribute list
  • Clean legacy HTML pages and convert them to Markdown while preserving headings, lists, links, and images

Best For

  • RAG engineers: turn public documentation pages into clean Markdown and JSON
  • Data analysts: batch fetch multiple URLs and summarize success, failure, and timing
  • Frontend or crawler engineers: extract target regions using class or id selectors
  • Data journalists: clean news pages and convert them into editable Markdown