skills/jeremylongshore/tons-of-skills-marketplace/creating-kubernetes-deployments/Gen Agent Trust Hub
creating-kubernetes-deployments
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides standard Kubernetes manifest templates and bash scripts that act as wrappers for kubectl commands.
- [COMMAND_EXECUTION]: The shell scripts (apply_manifest.sh, delete_manifest.sh, validate_manifest.sh, and get_resource_status.sh) utilize kubectl to interact with clusters. These operations are explicitly authorized in the skill's frontmatter under allowed-tools. The scripts implement safety measures including dry-run validations, confirmation prompts for destructive actions, and specific resource type iteration to ensure stability.
- [CREDENTIALS_UNSAFE]: No hardcoded credentials or sensitive tokens were identified. All configuration templates (e.g., secret_template.yaml, configmap_template.yaml) and instructional examples use generic placeholders such as "YOUR_VALUE_HERE" or benign example strings like "api-key-here".
- [EXTERNAL_DOWNLOADS]: No unauthorized or suspicious external downloads were detected. The skill references well-known official documentation (kubernetes.io) and the author's legitimate vendor resources (jeremylongshore.com, intentsolutions.io) for informational purposes.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided deployment requirements but focuses on generating structured YAML manifests and executing targeted kubectl commands. It provides extensive security guidelines to mitigate risks associated with untrusted container images or improper configurations.
Audit Metadata