az-aks-agent
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches configuration templates and installation scripts from official Microsoft and Azure sources (e.g., aka.ms and the Azure GitHub organization). These sources are well-known and legitimate for the skill's purpose.
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to execute Azure CLI (
az) and Kubernetes (kubectl) commands to analyze cluster health, retrieve logs, and monitor performance. - [PROMPT_INJECTION]: As a troubleshooting tool, the skill processes external data such as cluster events and container logs. This creates a potential surface for indirect prompt injection (Category 8), which is a common characteristic of diagnostic agents and is here limited to the cluster environment the user is already troubleshooting.
- Ingestion points: Cluster logs, events, and metrics retrieved via
kubectlandaz aks agentqueries. - Boundary markers: The skill instructions do not explicitly define delimiters for external data.
- Capability inventory: The agent can execute a variety of shell commands and cluster queries as defined in the documentation.
- Sanitization: No explicit data sanitization is described in the provided markdown instructions.
Audit Metadata