This is a continuation run because your previous response reached the model's output-token limit. Continue from exactly where it stopped without repeating completed content, and finish the pending work. If any fact, decision, or remaining action in the current thread is uncertain, use the available tools to retrieve the recent message history for that exact thread before continuing; do not guess.