There will be two inference responses after users stopping editing.

[10901008-RoryHuang]

After users stopping editing, InlineCompletion is triggered (triggerKind-Automatic), allowing a request to be sent to llm-ls. However, the response from llm-ls triggers InlineCompletion (triggerKind-Invoked ) again.
I'm a bit confused about this behavior. Could you please provide some more context?

[github-actions]

This issue is stale because it has been open for 30 days with no activity.