Headlamp AI plugin does not return pod logs

[sekar-anish]

I've installed the AI Plugin and have configured it to use an OpenAI deployment.

Version of Headlamp docker image: v0.35.0
Version of AI Plugin : version: 0.1.0-alpha

Context: I am trying to use the chat interface to analyze the logs of a pod and have the AI summarize the logs.

Issue: Instead of an analysis and a response, the chat response says it's starting to send a request to fetch logs and it just stops with that. Attaching a screenshot showing the interaction.

Expected outcome: Wanting to use the AI chat to read a container's logs, look through any errors and see if it can summarize what's going on. (possibly give remediations)

The chat stops after this.

[skoeva]

Hi, can you share what model you are using? and are you still experiencing this issue? I managed to get this:

[g0atcheez]

I get the same sort of output. It inserts the first line from the logs and then does nothing. I tried on both the default (gpt-4o) and gpt-4.1 with an OpenAI subscription.