Cut LLM token usage on command output and API traffic — compress what enters your coding agent's context, meter what your apps spend.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).