experiment-launch

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected.
  • [COMMAND_EXECUTION]: The skill uses a bundled helper script (gb-call) to interact with the GrowthBook REST API. All calls are limited to this specific tool, and input is structured via JSON payloads or specific API endpoints.
  • [CREDENTIALS_UNSAFE]: Credential management for GB_API_KEY follows standard practices, referencing environment variables or a local configuration file (~/.config/growthbook/.env). No hardcoded secrets are present.
  • [PROMPT_INJECTION]: The skill includes human-in-the-loop checkpoints (Step 6), requiring explicit user confirmation before experiments are started or configurations are published, which mitigates risks associated with automated execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 02:28 PM
Security Audit — agent-trust-hub — experiment-launch