Output just stops when model runs into sensitive content without explaining this

What version of Kimi Code is running?

0.17.1

Which open platform/subscription were you using?

kimi code

Which model were you using?

kimi-for-coding

What platform is your computer?

Linux 6.12.65-amd64-desktop-rolling x86_64 unknown

What issue are you seeing?

When the output gets to sensitive topic (Chinese gov/poltiics) the output just stops. and it keeps happening that I had to personally figure out what the problem is by noticing that it always stopped after or just before writing something related to poltiics. It's kind of ridiculous that the TUI doesn't explain what's wrong and it just stops mid-output. Surely the API is returning some info we can display that the censorship is the result for the stoppage so we can work around it. Instead I had to deduce what was going on.

What steps can reproduce the bug?

Tell it to write something political that would have it output the names of certain Chinese politicians.

What is the expected behavior?

Some error message or jus thte ouptut telling you that the topic is sensitive.

Additional information

No response

Start by reproducing the stop in the Kimi Code TUI with version 0.17.1, the kimi-for-coding model, and a political prompt. Trace how the API response is handled when generation ends mid-output; done means the TUI explains that sensitive content caused the stoppage instead of silently ending.