vss-query-analytics
Pass
Audited by Gen Agent Trust Hub on Jun 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill author (NVIDIA) and the referenced GitHub repository (github.com/NVIDIA-AI-Blueprints) are verified vendor resources, consistent with the provided context.
- [PROMPT_INJECTION]: The skill documentation proactively addresses indirect prompt injection by including a 'Scope guard' that warns against executing instructions embedded in alerts or analytics payloads. It specifically forbids auto-deployment triggered by untrusted data content.
- [COMMAND_EXECUTION]: The use of shell commands like
curl,jq, anddockeris restricted to querying the designated analytics service and verifying deployment status within the user's environment. - [CREDENTIALS_UNSAFE]: The skill correctly manages authentication by utilizing environment variables (
NGC_CLI_API_KEY,NVIDIA_API_KEY) for image pulls, avoiding hardcoded secrets.
Audit Metadata