Search the web and extract page content via the official Brave Search API — no browser required. Supports result counts, freshness filters, country targeting, and markdown content extraction.
---
name: brave-search
description: Web search and content extraction via the official Brave Search API. Use for searching docs, facts, or any web content — lightweight, no browser required.
---
Lightweight web research using the Brave Search API through bundled Node scripts.
## How to use
1. One-time setup: create a free Brave Search API account, export `BRAVE_API_KEY` in your shell profile, and run `npm install` in the skill directory.
2. Search with `search.js "query"`; add `-n 10` for more results (max 20), `--freshness pw` (or a `YYYY-MM-DDtoYYYY-MM-DD` range) to filter by time, and `--country DE` to localize.
3. Add `--content` to fetch and inline each result's page as markdown in one pass.
4. Extract a single page with `content.js <url>`, which returns readable markdown for citation or summarization.
Full skill & source: https://github.com/badlogic/pi-skills/tree/90bb51cae36515a648515b633a81c0c6efc8c74d/brave-searchSign in to rate and review this skill.
No reviews yet. Be the first to review this skill.