Motivation
Currently, once the agent starts a multi-step tool loop, users must wait for the entire execution to finish(if response.stop_reason != "tool_use") before providing feedback or corrections. This limits the ability to steer the agent when it's going off track or taking an incorrect approach.
Proposed Solution
A user can submit messages while the agent works. Enter sends a steering message (delivered after current tool, interrupts remaining tools). Alt + Enter sends a follow-up (waits until the agent finishes).
Motivation
Currently, once the agent starts a multi-step tool loop, users must wait for the entire execution to finish(
if response.stop_reason != "tool_use") before providing feedback or corrections. This limits the ability to steer the agent when it's going off track or taking an incorrect approach.Proposed Solution
A user can submit messages while the agent works. Enter sends a steering message (delivered after current tool, interrupts remaining tools). Alt + Enter sends a follow-up (waits until the agent finishes).