configuring-experiment-rollout

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional and defines how an agent should handle user requests for experiment configuration. It uses a structured approach to resolve ambiguity by asking the user for clarification before proceeding.
  • [EXTERNAL_DOWNLOADS]: The skill includes links to official PostHog documentation for further reading. These links target the vendor's own domain and do not involve executable code downloads or untrusted sources.
  • [SAFE]: No evidence of malicious patterns such as prompt injection, hardcoded credentials, obfuscation, or unauthorized data access was found. The instructions prioritize data integrity and user experience by requiring warnings for non-standard configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 04:13 AM