container-security

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches container management tools (k9s, ArgoCD, yq, Kustomize) from official GitHub release pages. These references target a well-known service and include version pinning to ensure reproducibility and integrity.
  • [COMMAND_EXECUTION]: Uses Azure CLI (az acr) and Trivy for container registry management and vulnerability scanning. These commands are necessary for the skill's intended purpose and operate within expected parameters.
  • [PROMPT_INJECTION]: The skill ingests and processes external JSON vulnerability reports from Trivy.
  • Ingestion points: Reads and parses trivy-results.json (SKILL.md).
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore embedded commands within the scan data.
  • Capability inventory: The skill has access to powerful tools including Bash, Write, Edit, and Agent (SKILL.md).
  • Sanitization: The provided Python parsing logic lacks sanitization or validation of the input fields beyond simple field extraction.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 01:13 PM
Security Audit — agent-trust-hub — container-security