Tavily
技能 已验证 活跃Tavily web search, content extraction, and research tools.
To provide a reliable and versatile tool for gathering information from the web, enabling users to conduct research, extract data from web pages, and obtain concise answers to queries.
功能
- Web search with advanced options
- Content extraction from URLs
- AI-generated answer summaries
- Domain filtering for searches
- Targeted content chunking
使用场景
- Conducting in-depth research on specific topics
- Extracting key information from multiple web pages
- Getting quick answers to factual questions
- Monitoring recent news or financial information
非目标
- Performing actions on web pages
- Interacting with user accounts or systems
- Providing real-time news alerts without a specific query
Invocation
- info:Overlapping near-synonym toolsWhile `web_search` and `tavily_search` serve similar purposes, the SKILL.md clearly delineates when to use each based on advanced options, minimizing ambiguity.
安装
npx skills add steipete/clawdis通过 npx 运行 Vercel skills CLI(skills.sh)— 需要本地安装 Node.js,以及至少一个兼容 skills 的智能体(Claude Code、Cursor、Codex 等)。前提是仓库遵循 agentskills.io 格式。
质量评分
已验证类似扩展
External Context
100Invoke parallel document-specialist agents for external web searches and documentation lookup
Web Search
100用于网络搜索。返回带有摘要、URL和缩略图的排名结果。支持新鲜度过滤、SafeSearch、用于自定义排名的Goggles以及分页。主搜索端点。
Exa Research Orchestrator
99Exa 赋能的深度研究。用于潜在客户开发、文献综述、深度分析、竞争对手分析，或任何单个搜索不足以满足需求的查询，包括诸如“研究这个”、“查找关于……的所有信息”、“找我所有的”或“深入研究……”之类的短语。
Deep Research
97Orchestrate multi-phase deep research with web search, memory retrieval, pattern matching, and synthesis into structured findings
Perplexity
95Web search and research using Perplexity AI. Use when user says "search", "find", "look up", "ask", "research", or "what's the latest" for generic queries. NOT for library/framework docs (use Context7) or workspace questions.
ARA Research Manager
100Records research provenance as a post-task epilogue, scanning conversation history at the end of a coding or research session to extract decisions, experiments, dead ends, claims, heuristics, and pivots, and writing them into the ara/ directory with user-vs-AI provenance tags. Use as a session epilogue — never during execution — to maintain a faithful, auditable trace of how a research project actually evolved.