Find why your Dynamo deployment failed and what to do next.
Diagnoses a Dynamo deployment failure and returns its problem class, strongest evidence, likely cause, and next corrective action.
When to use it
Use when a Dynamo deployment has failing pods, model-cache jobs, PVCs, workers, frontend/router health, endpoints, or benchmark jobs.
Give it the target namespace and deployment failure; it returns a diagnosis, evidence, and an exact next command or patch without executing remediation.
What you provide
Requires Python 3.10 or newer on the operator machine.
Requires kubectl configured for the target namespace, with permission to read pods, events, jobs, PVCs, and DynamoGraphDeployment resources.
This skill
http://127.0.0.1:8000/v1/chat/completions
Sends a test chat request
Kubernetes
Reads cluster status and logs