k8s-hpa-cost-tuning
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to run
kubectlcommands and local Node.js scripts to analyze cluster state and performance metrics. - [EXTERNAL_DOWNLOADS]: The utility scripts query Datadog's API endpoints to fetch metrics. These network requests target a well-known service for legitimate monitoring purposes.
- [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection (Category 8) due to the ingestion of untrusted data and the availability of sensitive capabilities.
- Ingestion points: The agent processes input from command-line arguments (cluster, namespace, and deployment names) and is designed to analyze production logs and HPA events.
- Boundary markers: The skill does not implement delimiters or specific instructions to prevent the agent from being influenced by malicious content embedded within the processed logs or metrics.
- Capability inventory: The agent has the ability to execute shell commands, perform outbound network requests via
fetch, and write files to the local disk. - Sanitization: Input arguments are used directly in the construction of Datadog queries and file paths without rigorous validation or escaping.
Audit Metadata