analytics-explore

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill interacts with the GrowthBook API using a local helper script and follows standard practices for handling API credentials via environment variables. It includes comprehensive logic for error handling and data interpretation.
  • [SAFE]: The skill's surface for indirect prompt injection was evaluated and determined to be safe. It processes structured data from the GrowthBook API, such as metric definitions and fact table schemas.
  • Ingestion points: Data enters through gb-call GET requests to GrowthBook API endpoints (e.g., /api/v1/fact-metrics, /api/v1/fact-tables).
  • Boundary markers: No explicit delimiters are specified for interpreting strings returned by the API.
  • Capability inventory: The skill is restricted to performing network requests via the gb-call tool and sleeping via the sleep command.
  • Sanitization: While no specific text sanitization is described, the skill's operations are limited to data reporting and chart rendering from a trusted vendor source.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 08:17 AM
Security Audit — agent-trust-hub — analytics-explore