The Agent Skills Directory

Indirect Prompt Injection Surface: The skill processes external Kubernetes manifests and cluster data to perform readiness assessments. This represents an indirect prompt injection surface; however, the skill mitigates this by enforcing read-only assessments, requiring explicit user approval for any proposed file changes, and providing clear boundary markers in its instructions.
Trusted External Resource Reference: The skill includes a reference to download and apply a YAML configuration for the Vertical Pod Autoscaler from the official Kubernetes project repository on GitHub. This is a well-known and trusted source for infrastructure components.
Credential Handling Safeguards: The instructions include specific guardrails to prevent the agent from requesting or outputting sensitive information such as access tokens, secret keys, or subscription identifiers, relying instead on local CLI authentication contexts.

azure-kubernetes