How is the search data formatted for the AI?

Results are returned as raw JSON specifically optimized for LLM consumption, featuring page titles, source URLs, and content snippets for accurate attribution.

Does the skill support live crawling?

Yes, you can use the --livecrawl parameter set to 'preferred' to ensure the skill fetches the most current version of a webpage rather than cached data.

Can I control the depth of the search results?

Yes, the skill supports three search types: 'fast' for quick lookups, 'auto' for balanced results, and 'deep' for comprehensive research queries.

Does this skill require an Exa AI API key?

No, it utilizes Exa AI's free MCP endpoint, allowing for immediate use without requiring an individual API key configuration.

What triggers the web search capability?

It is designed to trigger automatically on natural language prompts such as 'search the web', 'find online', or 'what's the latest on' specific topics.

Web Search

Name: Web Search
Author: vaayne

byvaayne

•

Web Scraping & Data Collection

Searches the web using Exa AI to provide real-time information retrieval and up-to-date data for AI coding workflows.

About

Web Search is a specialized Claude Code skill that bridges the gap between static training data and the live web by leveraging Exa AI's advanced search capabilities. It allows AI agents to perform real-time information retrieval, fetch current news, access the latest documentation, and conduct deep research with configurable search types like fast, auto, and deep. By returning LLM-optimized JSON results with source URLs and content snippets, it ensures that your coding environment stays informed with the most relevant and recent data available online.

Key Features

Advanced parameters for result count, timeouts, and live crawling
Real-time web search via Exa AI integration
21 GitHub stars
Configurable search depths including Fast, Auto, and Deep modes
LLM-optimized output with clean JSON snippets and source attribution
Zero-config setup utilizing free MCP endpoints

Use Cases

Fact-checking information or finding real-time news relevant to a coding project
Retrieving the latest software documentation or API changes beyond the model knowledge cutoff
Conducting market research or competitive analysis on recent technology trends

About

Key Features

Advanced parameters for result count, timeouts, and live crawling
Real-time web search via Exa AI integration
21 GitHub stars
Configurable search depths including Fast, Auto, and Deep modes
LLM-optimized output with clean JSON snippets and source attribution
Zero-config setup utilizing free MCP endpoints

Use Cases

Fact-checking information or finding real-time news relevant to a coding project
Retrieving the latest software documentation or API changes beyond the model knowledge cutoff
Conducting market research or competitive analysis on recent technology trends