Web Browsing
This skill enables browsing websites, extracting content from URLs, and searching the web for information.
When to Use
Use this skill when:
- User asks you to visit a website (e.g., "Check out example.com")
- User wants webpage content summarized
- User needs current information that requires internet search
- User provides a URL and wants it analyzed
- User asks to search for something online
How It Works
1. Direct URL Access
When given a specific URL:
User: "What's on https://example.com?"
→ Visit the page, extract main content, summarize key points
2. Web Search
When user asks to search:
User: "Find information about climate change"
→ Perform web search, present top results with summaries
3. Content Extraction
For specific data extraction:
User: "Get the latest news from techcrunch.com"
→ Navigate to site, extract relevant articles/headlines
Tools Available
- web_search: Search the web for information
- fetch_url: Visit and retrieve webpage content
- extract_content: Parse HTML and extract structured data
Best Practices
- Be specific - Tell me what you want from the page (summary, specific data, latest news)
- Provide URLs - If you have a specific page, share the URL directly
- Clarify intent - Let me know if you need:
- Quick summary
- Detailed analysis
- Specific data points
- Latest updates
示例
✅ "Visit https://news.ycombinator.com and summarize today's top stories"
✅ "Search for the latest React.js tutorial"
✅ "Check what's on Wikipedia's page about quantum computing"
✅ "Find pricing information from apple.com/iphone"
❌ Just say "browse the web" - be more specific!
Limitations
- Cannot interact with JavaScript-heavy sites (may miss dynamic content)
- Some sites block automated access
- Video/audio content cannot be played, only described if available
- Login-required pages won't work without credentials
Ready to browse! Just give me a URL or tell me what to search for. 🌐
依赖说明
运行环境
- Agent平台: 支持SKILL.md的任意AI Agent( Code / Cursor / Codex / CLI等)
- 操作系统: Windows / macOS / Linux
依赖说明
| 依赖项 | 类型 | 是否必需 | 获取方式 |
|---|---|---|---|
| LLM API | API | 必需 | 由Agent内置LLM提供 |
API Key 配置
- 本Skill基于Markdown指令,无需额外API Key(除内容中明确标注的外部API)
可用性分类
- 分类: MD+execute(纯Markdown指令,部分功能需要exec命令行执行能力)
- 说明: 基于Markdown的AI Skill,通过自然语言指令驱动Agent执行任务
核心能力
- Browse and summarize websites, extract content from URLs, search the web for information
- 触发关键词: web, websites, browsing, summarize, content, browse, extract
适用场景
| 场景 | 输入 | 输出 |
|---|---|---|
| 基础使用 | 用户请求 | 处理结果 |
不适用于：需要人工判断的复杂决策场景
使用流程
- 确认运行环境满足依赖说明中的要求
- 根据适用场景选择合适的使用方式
- 执行操作并检查输出结果
- 如遇错误，参考错误处理章节
错误处理
| 错误场景 | 原因 | 处理方式 |
|---|---|---|
| 配置错误 | 参数缺失或格式错误 | 检查依赖说明中的配置要求 |
| 运行时错误 | 运行环境不满足 | 确认运行环境符合依赖说明 |
| 网络错误 | 连接超时或不可达 | 检查网络连接后重试，参考国内替代方案 |
常见问题
Q1: 如何开始使用Web Browsing？
A: 请先阅读使用流程章节，确认环境满足依赖说明中的要求。
Q2: 遇到错误怎么办？
A: 请参考错误处理章节，按照表格中的处理方式操作。
Q3: Web Browsing有什么限制？
A: 请参考已知限制章节了解具体限制。