More token-efficient and focused Deep Search
Longer, more focused conversations with Deep Search address context window issues and enable even deeper research.
Deep Search now uses a dedicated subagent specialized for finding files relevant to your query. This subagent performs thorough searches and presents a summary to the main agent. Since the main Deep Search agent only needs to read this summary, this saves tokens in its context window.
This means Deep Search can explore more of your codebase while using fewer tokens, so you get more focused context and longer conversations - you should be able to ask Deep Search more questions before reaching the context window limit.
When Deep Search uses this subagent, you can follow along on the individual steps the subagent takes.
