gke-workload-scaling-troubleshooting
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [Standard Diagnostic Tooling]: The skill utilizes established command-line utilities like
kubectlandgcloudto gather diagnostic information from the cluster environment (SKILL.md). - [Read-Only Operation]: The instructions prioritize data retrieval and explicitly direct the agent to propose fixes for manual review rather than applying live changes to the infrastructure, which is a proactive safety measure (SKILL.md).
- [Trusted Documentation References]: All external links point to official technical documentation from Google Cloud and the Kubernetes project (SKILL.md).
- [Indirect Prompt Injection Surface]: The skill ingests and processes output from cluster diagnostic commands and logs, which constitutes a potential injection surface.
- Ingestion points: Data is retrieved from commands like
kubectl describe,kubectl get -o yaml, and Cloud Logging queries (SKILL.md). - Boundary markers: The skill establishes a "read-only diagnostics boundary" and requires presenting results as reviewable changes (SKILL.md).
- Capability inventory: The skill uses
kubectl,gcloud, andjqfor diagnostics throughout the resolution branches (SKILL.md). - Sanitization: No explicit content sanitization or filtering logic is specified for the raw data ingested from the cluster.
Audit Metadata