Can I customize the response style?

Yes, the '--system-prompt' flag allows you to guide the AI's answer style, such as requesting a response suitable for a specific audience or in a specific format.

How can I reduce token usage with this skill?

You can use the '--output-format toon' flag for a 40% token reduction or pipe JSON output to 'jq' to extract only the specific fields you need, which can reduce token usage by up to 80%.

Can I get structured data from this skill?

Yes, by using the '--output-schema' flag, you can define a custom JSON schema to ensure the AI returns structured, parseable data every time.

Why should I use 'answer' instead of 'search'?

Use 'answer' when you need a synthesized response that combines information from multiple sources. Use 'search' for simple lookups where you only need the URLs for manual verification.

What is the cost per answer with Exa Answer?

Each call to the Exa Answer skill costs approximately $0.005, making it a cost-effective solution for complex queries requiring AI synthesis.

Exa Answer

Name: Exa Answer
Author: benjaminjackson

bybenjaminjackson

•

ウェブスクレイピングとデータ収集

Generates fact-based answers and structured data from the web using AI-powered search and synthesis.

概要

Exa Answer is a high-performance Claude Code skill designed for tasks requiring factual accuracy and real-time web data. It leverages Exa.ai's neural search capabilities to synthesize answers with citations, extract specific information into custom JSON schemas, and optimize token usage through specialized output formats. This skill is ideal for developers needing to programmatically fetch verified facts, research technical topics, or build data-rich applications without manual browsing.

主な機能

Customizable system prompts for specific response styles
2 GitHub stars
Programmatic integration with jq support for field extraction
Structured data extraction using custom JSON schemas
AI-synthesized answers with direct web citations
Token-efficient 'toon' format for 40% lower costs

ユースケース

Generating summaries of complex topics for documentation and learning
Extracting structured product or company data from the live web
Automated technical research and fact-checking during development

概要

主な機能

Customizable system prompts for specific response styles
2 GitHub stars
Programmatic integration with jq support for field extraction
Structured data extraction using custom JSON schemas
AI-synthesized answers with direct web citations
Token-efficient 'toon' format for 40% lower costs

ユースケース

Generating summaries of complex topics for documentation and learning
Extracting structured product or company data from the live web
Automated technical research and fact-checking during development