Expand description
URL-fetch backend — bounded HTTP GET + HTML→text reduction.
Used by cortex::actions::ActionDispatcher to inline the body of
user-pasted URLs alongside web search results. The fetcher imposes
its own timeout and body cap so a runaway page can’t blow the
LLM context window or stall the answer.
Structs§
- Basic
UrlFetcher - Default HTTP fetcher: GET, follow redirects, strip HTML, return truncated text. No JavaScript execution, no auth, no cookies.