deploy-k8s-explorer

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). During Phase 5 the skill runs observe datastream view <DATASTREAM_ID> | wrap ... (and Phase 5a runs kubectl get pods ... | wrap ...), ingesting free-form JSON/metadata authored by the cluster/workloads into the assistant context.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly states it will read documentation pages from https://docs.observeinc.com at apply time to fetch YAML content and merge it into helm values (e.g., "https://docs.observeinc.com<docs_anchor>" and https://docs.observeinc.com/docs/kubernetes), so remote docs are fetched at runtime and directly control the agent's configuration.