Skills tavily-search
🔍

tavily-search

Safe

Search the Web with LLM-Optimized Results

Also available from: CK991357

Finding accurate, current information on the internet is time-consuming. This skill provides AI-ready search results optimized for LLM consumption with relevance scores, content snippets, and domain filtering.

Supports: Claude Codex Code(CC)
🥉 74 Bronze
1

Download the skill ZIP

2

Upload in Claude

Go to Settings → Capabilities → Skills → Upload skill

3

Toggle on and start using

Test it

Using "tavily-search". Search the web for 'AI coding assistants 2025'

Expected outcome:

Returns ranked results with title, URL, content snippet, and relevance score from recent sources

Using "tavily-search". Search for 'rust programming latest news' time-range week topic news

Expected outcome:

Returns recent news articles about Rust programming with dates and source credibility indicators

Security Audit

Safe
v1 • 4/23/2026

All 53 static findings are false positives. The skill is a legitimate web search wrapper for the Tavily CLI. External command patterns are documented CLI usage examples restricted to the `tvly` binary only. Network access is limited to the official Tavily API. No arbitrary command execution, credential exfiltration, or malicious patterns detected. Risk level: SAFE.

1
Files scanned
92
Lines analyzed
0
findings
1
Total audits
No security issues found
Audited by: claude

Quality Score

38
Architecture
100
Maintainability
85
Content
50
Community
100
Security
91
Spec Compliance

What You Can Build

Research and Fact-Checking

Quickly gather current information on any topic with relevance-ranked sources and content snippets for verification.

News Monitoring

Track recent developments on specific topics, companies, or technologies with time-filtered news results.

Technical Documentation Search

Find official documentation, API references, and technical articles from trusted domains.

Try These Prompts

Basic Web Search 
Search the web for {topic}
News Search with Date Filter 
Search for recent news about {topic} from the past week
Domain-Filtered Search 
Search for {topic} from official sources like example.com and developer.example.io
Advanced Research Search 
Search for {topic} using advanced depth, include full content and AI answers, max 10 results

Best Practices

  • Use specific, focused queries under 400 characters for better results
  • Filter by trusted domains when researching technical topics
  • Choose appropriate search depth: basic for general use, advanced for precision tasks

Avoid

  • Avoid vague queries like 'tell me about technology' - be specific about the topic
  • Do not use this for real-time chat autocomplete - use ultra-fast depth only when latency is critical
  • Avoid excluding all domains if you need authoritative results - use include-domains instead

Frequently Asked Questions

How do I get started with tavily-search?
Install the Tavily CLI with curl -fsSL https://cli.tavily.com/install.sh | bash, then run tvly login with your API key. The skill will then execute searches on your behalf.
What is the difference between search depths?
Ultra-fast is fastest with lower relevance, basic is the default balance, and advanced provides the highest relevance at slower speed for precision tasks.
Can I search for content from specific websites?
Yes, use --include-domains to restrict results to specific domains like --include-domains github.com,stackoverflow.com.
How do I get full article content in search results?
Add --include-raw-content to your search command to include full page content alongside the search results.
What happens if my search query is too long?
Keep queries under 400 characters. For complex research, break it into multiple focused sub-queries.
Is this skill suitable for real-time applications?
Use the ultra-fast depth for real-time needs like autocomplete. For research and fact-checking, use basic or advanced depth for better relevance.

Developer Details

Author

tavily-ai

License

MIT

Repository

https://github.com/tavily-ai/skills/tree/main/skills/tavily-search/

Ref

main

File structure

📄 SKILL.md