grafana-dashboard

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFECREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The references/docker-compose-setup.md file contains a default administrator password (admin) for the Grafana service defined in the GF_SECURITY_ADMIN_PASSWORD environment variable. While the SKILL.md documentation includes a specific warning to avoid leaving default passwords, the presence of the default in a working example poses a risk for insecure deployments.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and visualize data from external sources, which creates a surface for indirect prompt injection attacks.
  • Ingestion points: Metric labels, names, and time-series data from external Prometheus data sources, as referenced in references/grafana-dashboard-json.md and references/grafana-alert-configuration.md.
  • Boundary markers: No explicit delimiters or instructions to ignore instructions embedded in metric data are present.
  • Capability inventory: The skill includes a JavaScript API client (references/grafana-api-client.md) that can create or modify dashboards and alerts, and a shell script (scripts/health-check.sh) capable of performing network-based health checks.
  • Sanitization: There are no documented validation or sanitization routines to ensure that untrusted external metric strings are safely handled before interpolation into dashboard JSON or API calls.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 10:44 AM
Security Audit — agent-trust-hub — grafana-dashboard