Skip to content
Kitbay

llm-context

MITUpdated on Nov 9, 2026196GitHub stars

A Brave skill for the Brave Search API's LLM context endpoint, which returns pre-extracted text, tables and code from web pages for retrieval-augmented generation rather than a list of links.

What the skill teaches

A Brave skill for the Brave Search API's LLM context endpoint, which returns pre-extracted text, tables and code from web pages for retrieval-augmented generation rather than a list of links. It explains how this differs from the answers endpoint, how to call it by GET or POST, how to size the context with token and result limits, threshold modes, local and place recall with location headers, and Goggles for custom ranking, along with the response format. It needs a Brave Search API key set as BRAVE_SEARCH_API_KEY.

Capabilities

  • Fetch extracted web content for RAG grounding
  • Control context size by tokens and result count
  • Add location context to searches
  • Re-rank sources with Goggles