aks-troubleshooting

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • Privileged Debugging Capabilities: The skill utilizes Inspektor Gadget for deep diagnostics, which involves launching a privileged debug pod using kubectl debug --profile=sysadmin. This is a powerful administrative operation necessary for low-level node observability. The skill includes appropriate guardrails by explicitly requiring user approval and highlighting the read-only nature of the gadgets used.
  • Automated Diagnostic Scripts: Included scripts like cluster-snapshot.sh and pod-deep-dive.sh automate the collection of cluster and pod state. These scripts use standard kubectl and az commands. They use POSIX shell best practices, such as variable quoting, to handle cluster identifiers safely.
  • External Resource Downloads: The skill references container images from the Microsoft Container Registry (mcr.microsoft.com) to run diagnostic tools. These resources originate from a well-known service associated with the vendor and are used within the scope of the skill's troubleshooting purpose.
  • Indirect Prompt Injection Surface: The skill processes dynamic data from the Kubernetes environment, such as pod names, node names, and log content. While this creates a surface for indirect prompt injection if malicious content is present in logs or metadata, the skill focuses on evidence-based reporting and does not include autonomous write actions by default.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 08:18 AM
Security Audit — agent-trust-hub — aks-troubleshooting