[BUG] UR 🤖 gets lost and leaves the human at an empty input prompt. Telling ur 🤖 "continue" seems to work.

Environment

• Platform (select one):
• [√] Anthropic API
• [ ] AWS Bedrock
• [ ] Google Vertex AI
• [ ] Other: <!-- specify -->
• Claude CLI version: 0.2.105 (Claude Code)
• Operating System: macOS 15.3
• Terminal: IntelJIdea

Bug Description

I don't have a "clear and concise" description here, I asked the model to look at an issue. The model made a bunch of code changes, and returned - leaving the human looking at an empty input line. Not knowing WTF the model wanted, I prompted "continue". The model was apparently finished with code changes and continued to write a .md file then report that it was done. This has happened similarly before, it's fairly infrequent.

!Image

Steps to Reproduce

1. Yeah , about that...

Expected Behavior

The model to finish the task, or to otherwise provide some response.

Actual Behavior

The model apparently finished the coding task, but did not return a summary/response.

Additional Context

I would like to know -what- could I report here that would be more useful than this vague hand-wavey report? Is there more data I could have pulled to attach to this (and how)?