Tavily Plus

面向智能网络搜索的 Skill：识别查询意图，将复杂问题拆解为子问题，按 research/quick/compare 模式路由搜索深度，遇到限流时轮换多个 Tavily 密钥，汇总结果生成摘要，并可导出带时间戳的 Markdown 报告。提供 smart_search、tavily_search、tavily_extract 三个工具，依赖 Tavily 服务。

fundou1081

热度 31.85

最近更新 2026-05-11 17:39:08

分类：搜索依赖：代码执行 / 第三方 API / 读写文件 / 内部调用大模型应用场景：研究 / 摘要总结 / 报告撰写 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add tavily-plus

Metadata

name: tavily-plus
description: Enhanced Tavily search with multi-API key rotation, AI-powered intent recognition, sub-question decomposition, intelligent summarization, and offline document export. Use when user needs intelligent web search with auto-failover, research workflows, or downloadable reports.
version: 1.0.0
metadata: {"openclaw":{"emoji":"🔍","requires":{"config":["plugins.entries.tavily.config"]}}}

Tavily Plus

Enhanced search skill with AI augmentation and multiple API keys.

Tools

Tool	Description
`smart_search`	AI-enhanced search with intent recognition, sub-question breakdown, and smart routing
`tavily_search`	Standard Tavily search (fallback/legacy mode)
`tavily_extract`	URL content extraction

Configuration

Configure in plugins.entries.tavily.config.webSearch:

{
  "apiKey": {
    "key1": "tvly-xxx1",
    "key2": "tvly-xxx2"  
  },
  "baseUrl": "https://api.tavily.com"
}

Or via env: TAVILY_API_KEY, TAVILY_API_KEY_2, TAVILY_API_KEY_N

smart_search

AI-augmented search that performs:

Intent Recognition - Classify query type (factual, news, research, comparison, how-to, etc.)
Sub-question Decomposition - Break complex queries into atomic sub-queries
Multi-key Rotation - Rotate through configured API keys on rate limit
Smart Search Execution - Route each sub-query to optimal depth/topic
Intelligent Summarization - Synthesize findings into coherent answer
Offline Document Export - Optional markdown/PDF report generation

Parameters

Parameter	Type	Description
`query`	string	Natural language search query
`mode`	string	`research` (deep), `quick` (basic), `compare` (multiple)
`max_results`	number	Results per sub-query (1-20)
`include_answer`	boolean	Include AI summary (default: true)
`export_doc`	boolean	Generate offline markdown report
`sub_questions`	array	Optional pre-defined sub-queries (bypasses decomposition)

Query Modes

Mode	Behavior
`research`	Deep search with multiple sub-queries, high max_results
`quick`	Single-pass basic search
`compare`	Multi-perspective search for comparisons

Response Format

{
  "intent": "research",
  "sub_questions": ["...", "..."],
  "search_results": [...],
  "summary": "...",
  "doc_path": "/path/to/report.md"  // if export_doc=true
}

Key Rotation Logic

Check plugins.entries.tavily.config.webSearch.apiKey (object with multiple keys)
Fall back to env vars: TAVILY_API_KEY, TAVILY_API_KEY_2, ..., TAVILY_API_KEY_N
On 429 (rate limit), automatically rotate to next key
Log which key was used in response metadata

Intent Classification

Supported intent types:

factual - Direct answer queries (who, what, when, where)
how_to - Procedural instructions
comparison - A vs B analysis
research - In-depth information gathering
news - Current events
opinion - Perspectives and opinions

Offline Export

When export_doc: true:

Generate timestamped markdown report
Include: intent, sub-questions, each search result with source, AI summary
Save to ~/.openclaw/workspace/reports/tavily-plus-{timestamp}.md
Return doc_path in response

Workflow Summary

User Query
    ↓
Intent Recognition (LLM)
    ↓
Sub-question Decomposition (LLM)
    ↓
For each sub-question:
    ├─ Query Tavily (key rotation on 429)
    └─ Collect results
    ↓
Intelligent Summarization (LLM)
    ↓
[Optional] Generate Offline Report
    ↓
Return Enhanced Response