bert-builder

Tavily AI Search

知识管理v1.0.0
作者:bert-builder更新于 2026-09-16tavily@clawhub_bert-builder/tavily
信息检索摘要总结#Web Search#News

据 AlphaMoat 统计,截至 2026-10,Tavily AI Search 累计下载 103,117 次、获 43 个星标,按累计下载量在本次参与排名的 10,914 个 AI 技能中排名第 240 位,在「知识管理」分类(745 个)中排名第 21 位。

列表默认按累计下载量排序,也可按星标排序;详情页排名按累计下载量计算。

累计下载
10.3万
星标
43
全站下载量排名
#240
全站前 3% · 「知识管理」前 3%
平台累计下载星标评论
ClawHub ↗4.9万36-
SkillHub ↗10.3万43—

各平台的计数口径不同,统计时间不一致。

这个技能是做什么的

技能面向大模型使用场景,返回结构化、易解析的检索结果,并可选生成答案摘要,另支持域名过滤、新闻检索、图片搜索与原文抓取。检索分 basic 与 advanced 两档,深度检索应按需升级而不是默认使用。 查询会传送至 Tavily;避免输入私人敏感内容,并依回传的来源页核验摘要。

资料发布于 2026-10-29

可以用来做什么

  • 为研究型问题收集带来源的检索结果
  • 限定站点范围检索新闻或做事实核对

什么时候用

  • 需要干净、便于模型消费的搜索结果时
  • 要按域名收窄检索范围时

示例提示词

以下是根据技能文档生成的示例,不代表实际运行结果。

帮我检索某个行业最近的监管变化,只要指定几个站点的结果,并附来源
这个问题先跑基础检索,如果覆盖不够再说明为什么要升级到深度模式

月度数据趋势

累计下载10.3万
3.7万7.2万10.8万2026-05 · 累计下载 4.1万2026-06 · 累计下载 4.3万2026-07 · 累计下载 4.5万2026-08 · 累计下载 5.6万2026-09 · 累计下载 6.3万2026-10 · 累计下载 10.3万26-0526-0926-10
星标43
3841432026-05 · 星标 382026-06 · 星标 402026-07 · 星标 432026-08 · 星标 432026-09 · 星标 432026-10 · 星标 4326-0526-0926-10

2026-09 起统计方法调整:同名技能按作者分别统计,并覆盖全部技能。图中虚线为调整前的月份,与之后的数据不直接比较,也不计算跨越调整的增长。

月份累计下载环比星标
2026-1010.3万—43
2026-096.3万—43
2026-085.6万+24.3%43
2026-074.5万+6.0%43
2026-064.3万+3.4%40
2026-054.1万—38

SKILL.md 原文

以下是作者发布的原始文件,按纯文本显示,本站未做修改或审核;其中的链接不可点击,命令请核实后再运行。抓取于 2026-10-23。

9.8 KB · ClawHub v1.0.0
---
name: tavily
description: AI-optimized web search using Tavily Search API. Use when you need comprehensive web research, current events lookup, domain-specific search, or AI-generated answer summaries. Tavily is optimized for LLM consumption with clean structured results, answer generation, and raw content extraction. Best for research tasks, news queries, fact-checking, and gathering authoritative sources.
---

# Tavily AI Search

## Overview

Tavily is a search engine specifically optimized for Large Language Models and AI applications. Unlike traditional search APIs, Tavily provides AI-ready results with optional answer generation, clean content extraction, and domain filtering capabilities.
展开全部(共 415 行)

**Key capabilities:**
- AI-generated answer summaries from search results
- Clean, structured results optimized for LLM processing
- Fast (`basic`) and comprehensive (`advanced`) search modes
- Domain filtering (include/exclude specific sources)
- News-focused search for current events
- Image search with relevant visual content
- Raw content extraction for deeper analysis

## Architecture

```mermaid
graph TB
    A[User Query] --> B{Search Mode}
    B -->|basic| C[Fast Search<br/>1-2s response]
    B -->|advanced| D[Comprehensive Search<br/>5-10s response]
    
    C --> E[Tavily API]
    D --> E
    
    E --> F{Topic Filter}
    F -->|general| G[Broad Web Search]
    F -->|news| H[News Sources<br/>Last 7 days]
    
    G --> I[Domain Filtering]
    H --> I
    
    I --> J{Include Domains?}
    J -->|yes| K[Filter to Specific Domains]
    J -->|no| L{Exclude Domains?}
    K --> M[Search Results]
    L -->|yes| N[Remove Unwanted Domains]
    L -->|no| M
    N --> M
    
    M --> O{Response Options}
    O --> P[AI Answer<br/>Summary]
    O --> Q[Structured Results<br/>Title, URL, Content, Score]
    O --> R[Images<br/>if requested]
    O --> S[Raw HTML Content<br/>if requested]
    
    P --> T[Return to Agent]
    Q --> T
    R --> T
    S --> T
    
    style E fill:#4A90E2
    style P fill:#7ED321
    style Q fill:#7ED321
    style R fill:#F5A623
    style S fill:#F5A623
```

## Quick Start

### Basic Search

```bash
# Simple query with AI answer
scripts/tavily_search.py "What is quantum computing?"

# Multiple results
scripts/tavily_search.py "Python best practices" --max-results 10
```

### Advanced Search

```bash
# Comprehensive research mode
scripts/tavily_search.py "Climate change solutions" --depth advanced

# News-focused search
scripts/tavily_search.py "AI developments 2026" --topic news
```

### Domain Filtering

```bash
# Search only trusted domains
scripts/tavily_search.py "Python tutorials" \
  --include-domains python.org docs.python.org realpython.com

# Exclude low-quality sources
scripts/tavily_search.py "How to code" \
  --exclude-domains w3schools.com geeksforgeeks.org
```

### With Images

```bash
# Include relevant images
scripts/tavily_search.py "Eiffel Tower architecture" --images
```

## Search Modes

### Basic vs Advanced

| Mode | Speed | Coverage | Use Case |
|------|-------|----------|----------|
| **basic** | 1-2s | Good | Quick facts, simple queries |
| **advanced** | 5-10s | Excellent | Research, complex topics, comprehensive analysis |

**Decision tree:**
1. Need a quick fact or definition? → Use `basic`
2. Researching a complex topic? → Use `advanced`
3. Need multiple perspectives? → Use `advanced`
4. Time-sensitive query? → Use `basic`

### General vs News

| Topic | Time Range | Sources | Use Case |
|-------|------------|---------|----------|
| **general** | All time | Broad web | Evergreen content, tutorials, documentation |
| **news** | Last 7 days | News sites | Current events, recent developments, breaking news |

**Decision tree:**
1. Query contains "latest", "recent", "current", "today"? → Use `news`
2. Looking for historical or evergreen content? → Use `general`
3. Need up-to-date information? → Use `news`

## API Key Setup

### Option 1: Clawdbot Config (Recommended)

Add to your Clawdbot config:

```json
{
  "skills": {
    "entries": {
      "tavily": {
        "enabled": true,
        "apiKey": "tvly-YOUR_API_KEY_HERE"
      }
    }
  }
}
```

Access in scripts via Clawdbot's config system.

### Option 2: Environment Variable

```bash
export TAVILY_API_KEY="tvly-YOUR_API_KEY_HERE"
```

Add to `~/.clawdbot/.env` or your shell profile.

### Getting an API Key

1. Visit https://tavily.com
2. Sign up for an account
3. Navigate to your dashboard
4. Generate an API key (starts with `tvly-`)
5. Note your plan's rate limits and credit allocation

## Common Use Cases

### 1. Research & Fact-Finding

```bash
# Comprehensive research with answer
scripts/tavily_search.py "Explain quantum entanglement" --depth advanced

# Multiple authoritative sources
scripts/tavily_search.py "Best practices for REST API design" \
  --max-results 10 \
  --include-domains github.com microsoft.com google.com
```

### 2. Current Events

```bash
# Latest news
scripts/tavily_search.py "AI policy updates" --topic news

# Recent developments in a field
scripts/tavily_search.py "quantum computing breakthroughs" \
  --topic news \
  --depth advanced
```

### 3. Domain-Specific Research

```bash
# Academic sources only
scripts/tavily_search.py "machine learning algorithms" \
  --include-domains arxiv.org scholar.google.com ieee.org

# Technical documentation
scripts/tavily_search.py "React hooks guide" \
  --include-domains react.dev
```

### 4. Visual Research

```bash
# Gather visual references
scripts/tavily_search.py "modern web design trends" \
  --images \
  --max-results 10
```

### 5. Content Extraction

```bash
# Get raw HTML content for deeper analysis
scripts/tavily_search.py "Python async/await" \
  --raw-content \
  --max-results 5
```

## Response Handling

### AI Answer

The AI-generated answer provides a concise summary synthesized from search results:

```python
{
  "answer": "Quantum computing is a type of computing that uses quantum-mechanical phenomena..."
}
```

**Use when:**
- Need a quick summary
- Want synthesized information from multiple sources
- Looking for a direct answer to a question

**Skip when** (`--no-answer`):
- Only need source URLs
- Want to form your own synthesis
- Conserving API credits

### Structured Results

Each result includes:
- `title`: Page title
- `url`: Source URL
- `content`: Extracted text snippet
- `score`: Relevance score (0-1)
- `raw_content`: Full HTML (if `--raw-content` enabled)

### Images

When `--images` is enabled, returns URLs of relevant images found during search.

## Best Practices

### 1. Choose the Right Search Depth

- Start with `basic` for most queries (faster, cheaper)
- Escalate to `advanced` only when:
  - Initial results are insufficient
  - Topic is complex or nuanced
  - Need comprehensive coverage

### 2. Use Domain Filtering Strategically

**Include domains for:**
- Academic research (`.edu` domains)
- Official documentation (official project sites)
- Trusted news sources
- Known authoritative sources

**Exclude domains for:**
- Known low-quality content farms
- Irrelevant content types (Pinterest for non-visual queries)
- Sites with paywalls or access restrictions

### 3. Optimize for Cost

- Use `basic` depth as default
- Limit `max_results` to what you'll actually use
- Disable `include_raw_content` unless needed
- Cache results locally for repeated queries

### 4. Handle Errors Gracefully

The script provides helpful error messages:

```bash
# Missing API key
Error: Tavily API key required
Setup: Set TAVILY_API_KEY environment variable or pass --api-key

# Package not installed
Error: tavily-python package not installed
To install: pip install tavily-python
```

## Integration Patterns

### Programmatic Usage

```python
from tavily_search import search

result = search(
    query="What is machine learning?",
    api_key="tvly-...",
    search_depth="advanced",
    max_results=10
)

if result.get("success"):
    print(result["answer"])
    for item in result["results"]:
        print(f"{item['title']}: {item['url']}")
```

### JSON Output for Parsing

```bash
scripts/tavily_search.py "Python tutorials" --json > results.json
```

### Chaining with Other Tools

```bash
# Search and extract content
scripts/tavily_search.py "React documentation" --json | \
  jq -r '.results[].url' | \
  xargs -I {} curl -s {}
```

## Comparison with Other Search APIs

**vs Brave Search:**
- ✅ AI answer generation
- ✅ Raw content extraction
- ✅ Better domain filtering
- ❌ Slower than Brave
- ❌ Costs credits

**vs Perplexity:**
- ✅ More control over sources
- ✅ Raw content available
- ✅ Dedicated news mode
- ≈ Similar answer quality
- ≈ Similar speed

**vs Google Custom Search:**
- ✅ LLM-optimized results
- ✅ Answer generation
- ✅ Simpler API
- ❌ Smaller index
- ≈ Similar cost structure

## Troubleshooting

### Script Won't Run

```bash
# Make executable
chmod +x scripts/tavily_search.py

# Check Python version (requires 3.6+)
python3 --version

# Install dependencies
pip install tavily-python
```

### API Key Issues

```bash
# Verify API key format (should start with tvly-)
echo $TAVILY_API_KEY

# Test with explicit key
scripts/tavily_search.py "test" --api-key "tvly-..."
```

### Rate Limit Errors

- Check your plan's credit allocation at https://tavily.com
- Reduce `max_results` to conserve credits
- Use `basic` depth instead of `advanced`
- Implement local caching for repeated queries

## Resources

See [api-reference.md](references/api-reference.md) for:
- Complete API parameter documentation
- Response format specifications
- Error handling details
- Cost and rate limit information
- Advanced usage examples

## Dependencies

- Python 3.6+
- `tavily-python` package (install: `pip install tavily-python`)
- Valid Tavily API key

## Credits & Attribution

- Tavily API: https://tavily.com
- Python SDK: https://github.com/tavily-ai/tavily-python
- Documentation: https://docs.tavily.com

「知识管理」分类下的热门技能

搜索引擎知识管理

通过公开搜索URL在16个搜索引擎(7个国内直连+9个国际)中做跨引擎检索、结果去重与多源交叉验证,无需API Key,内置零依赖Python脚本。Use when 用户要求“搜一下/查资料/多找几个来源/核实消息/找公众号文章/排查报错/找论文或开源项目”时。输入自然语言查询词,输出带来源链接与可信度标注的结构化结果。不适用于私有数据检索、批量抓取、需要登录后可见的内容。

↓ 1075.9万★ 147u_8ee34d77
ima-skills知识管理

通过 IMA OpenAPI 管理笔记与知识库内容。

↓ 108.4万★ 566u_1742680f
Summarize知识管理

把长文本、文档与网页内容压缩成要点和关键词的摘要工具。

↓ 87.6万★ 1,048paudyyin
全网搜索知识管理

通过“查一下、搜一下、搜索、找一下”等自然语言检索公开网页,整理来源并交付横版 PDF、Markdown 与标题链接表。支持用户主动开启、查看和取消技能收藏,保留原报告取回。

↓ 50.9万★ 2u_32d99f8a
web-tools-guide知识管理

规定如何为网页请求选择合适工具的必读指引。

↓ 40.0万★ 220user_ec205dbb
Aihot Skill Lite知识管理

AI HOT(aihot.virxact.com)中文 AI 资讯查询 Skill,帮助用户获取今日 AI 圈动态、AI 日报、AI 热点、最新 AI 资讯以及 OpenAI、Anthropic、Google 等公司的最新发布信息。

↓ 32.1万★ 92Khazix
全球12亿文献知识库(8千万中文期刊可下载)知识管理

全球12亿文献知识库(8千万中文期刊可下载)——通过 SmartLib 开放平台 API 提供中外文学术文献检索与下载能力,覆盖 8000 万篇授权中文期刊全文 + 12.28 亿条全球文献元数据(期刊 7.19 亿 / 专利 2.15 亿 / 会议 7155 万 / 学位论文 2473 万 / 标准 268 万)。 能力:中英文关键词检索、文献详情、中文期刊 PDF 全文下载、外文 OA 文献十级渠道免费下载(不消耗配额)、智能关键词扩展、核心期刊优先排序、相关性重排、引文追溯、分类号检索。 配额:首次使用自动注册,免费 100 次检索 + 10 次下载 / 月;耗尽自动弹出套餐(体验卡 / 个人版月 / 专业版月 / 单篇下载 / 下载包),可说「升级 / 充值」唤起;企业 / 机构定制联系 [email protected]。 触发:用户表达"查论文""找文献""检索学术""搜期刊""查专利""找标准""下论文""写文献综述""找参考文献""查 SCI/EI"等意图时启用;也适用"帮我找关于 XX 的论文""写文献综述""找几篇引用支撑论点"。英文:"find papers" "search literature" "write literature review" "find supporting citations"。 调用前必须先用 /consume 获取 consume_token,再凭 token 调 /search(每次计费接口调用都需一次 /consume)。

↓ 30.3万★ 189user_164f4c1f
Markdown Converter知识管理

用 uvx 直接调用 markitdown,把各类文件转成 Markdown。

↓ 20.4万★ 256Peter Steinberger

数据月份:2026-10 · 下载/星标数聚合自公开技能注册库(ClawHub、SkillHub),按月更新。口径说明见数据方法。