Tommy Nguyen
Tommy Nguyen
@droot ✅ Curious if we can configure the CI to auto-run so that contributors can catch CI bugs automatically?
@Keyruu the https://github.com/mark3labs/mcp-go was upgraded to version 0.31.0 indirectly via https://github.com/GoogleCloudPlatform/kubectl-ai/pull/291/commits/bcfb09556c254868843c695182ecc3599d5d4aeb . Could you check if it fix your issue by any chance?
I think this is expected behavior, and trying to mutate the default output of `kubectl logs` could lead to confusion. Instead, prompting the LLM to use an MCP server like...
👍🏼 for whitelist / blacklist pattern
No worry. As long as we have a straight-forward and conventional way to set flag, we're in good shape.
Great, thanks @droot. Then I'll continue the on-going PRs with non-viper implementation.
https://github.com/GoogleCloudPlatform/kubectl-ai/issues/210
> For some reason, this model sometimes get stuck into returning simply kubectl() as a response, still, it appears to be something specific for this model I have noticed similar...
``` λ google-kubectl-ai % ./kubectl-ai --llm-provider openai --model gpt-4.1 Hey there, what can I help you with today? >>> hello Hello! 👋 How can I assist you with your Kubernetes...
@zvdy Somehow for my testing with GPT 4.1, it doesn't work at all. And then it just tried to call `kubectl` without any action.