GLM 5.2 not functional

[eternaldensity]

I was using OpenCode Go, which functioned with other models.
The log doesn't specify what tool call it was attempting to make and fail.
tui-4200.log

Here's what was going on in the TUI (filenames shortened)
openclaudia.txt
It seems it did successfully read the files, but it showed no chain of thought and no output other that the read results. Perhaps OpenClaudia isn't understanding most of the response format?

[eternaldensity]

same problem with Kimi models