skills/grafana/gcx/gcx-observability/Gen Agent Trust Hub

gcx-observability

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill reads and processes local application files, which constitutes a surface for indirect prompt injection if those files contain malicious instructions meant to subvert the agent's behavior.
  • Ingestion points: The skill scans project Dockerfiles in Phase 3 to identify instrumentation compatibility issues and reads existing YAML and JavaScript configuration files for SLOs, synthetic checks, k6 tests, and dashboards across various phases.
  • Boundary markers: The instructions lack specific delimiters or negative constraints to prevent the agent from interpreting content within these files as instructions.
  • Capability inventory: The skill is granted high-privilege tools including Bash, kubectl, and gcx, which are used to execute the deployment and configuration tasks derived from the ingested data.
  • Sanitization: No explicit sanitization or filtering of the file content is performed before the agent acts upon the discovered information.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 07:09 AM
Security Audit — agent-trust-hub — gcx-observability