deploying-monitoring-stacks
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONMETADATA_POISONINGEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates untrusted user input into generated deployment manifests and configuration files.
- Ingestion points: User-provided configuration parameters (e.g., CPU thresholds, Slack webhooks) as described in
SKILL.md. - Boundary markers: Absent. The instructions do not define clear delimiters for the agent to ignore instructions embedded in user-supplied data.
- Capability inventory: The skill uses Python helper scripts to perform file system writes and has access to
Bashfor deployment commands. - Sanitization: Absent. The scripts perform direct file operations without validating the content of user-supplied data.
- [EXTERNAL_DOWNLOADS]: The skill recommends installing official monitoring charts.
- Source: https://github.com/prometheus-community/helm-charts/tree/main/charts/kube-prometheus-stack via Helm.
- Method: helm install.
- [METADATA_POISONING]: A documentation discrepancy exists in
scripts/README.md, which lists scripts with .sh extensions, while the repository contains Python scripts with .py extensions.
Audit Metadata