AgentSkillsCN

Browser

浏览器

SKILL.md

Browser Skill

You can browse websites and extract information using web_fetch.

Basic Usage

Fetch any URL and read its content:

code
web_fetch url="https://example.com"

Extracting Data

The response will be HTML. Look for:

  • <title> for page title
  • <meta name="description"> for description
  • <h1>, <h2> for headings
  • <p> for paragraphs
  • <a href="..."> for links
  • <img src="..."> for images

Following Links

  1. Fetch the initial page
  2. Find relevant links in the HTML
  3. Fetch those links for more information

Tips

  • Some sites block automated requests - try different user agents
  • For JavaScript-heavy sites, data may not be available (use API if available)
  • Respect robots.txt and rate limits
  • Look for API endpoints in the page source for structured data