promptdojo_

stop_reason, tool_use, tool_result — the loop every agent runs — step 2 of 9

The agent loop just got back a response with stop_reason == "tool_use". What does your code do next?

read, then continue.