brave-search

To perform headless web search and extract content to Markdown, run the search script with a query using Node.js. It fetches Brave Search results and applies Readability and Turndown to convert web pages into structured Markdown text.

Yes, you need a Brave API key to fetch web search results programmatically. You must configure the BRAVE_API_KEY environment variable to authenticate your requests and retrieve structured search results.

Yes, you can configure the number of web search results returned by the Brave Search API. The script supports configurable result counts so you can limit or expand the titles, links, and snippets retrieved for your specific query.

The best way to extract readable content for AI assistants is using Readability and Turndown. These libraries strip clutter from web pages fetched via Brave Search and convert the remaining text into clean Markdown for offline reading or summarization.

Yes, the headless search approach supports GitHub Flavored Markdown by using the turndown-plugin-gfm dependency. It ensures extracted web content is properly formatted with GFM features like tables and task lists during content conversion.

Using Readability and jsdom for web content extraction isolates the main article text from website boilerplate. This provides a clean DOM structure for Turndown to convert into readable Markdown, avoiding noise from raw HTML parsing.