Brave Search

Headless web search and content extraction using Brave Search. No browser required.

Setup

Assume packages are already installed. If not, run once:

cd ~/.pi/agent/skills/brave-search/ && npm ci

Search

# Basic search (5 results)
~/.pi/agent/skills/brave-search/search.js "query"
# More results
~/.pi/agent/skills/brave-search/search.js "query" -n 10

Extract Page Content

# Not part of this skill, but useful in conjunction with it
read_web_page <url>

Fetches a URL and extracts readable content as markdown.

When to Use

• Searching for documentation or API references
• Looking up facts or current information
• Fetching content from specific URLs
• Any task requiring web search without interactive browsing