finding-experiments
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were detected.
- [SAFE]: The skill correctly instructs the agent to use dedicated PostHog experiment tools for identification rather than generic feature flag tools, ensuring accurate data retrieval.
- [SAFE]: The use of the Posthog-local MCP server represents standard, legitimate integration with vendor-provided infrastructure for the author 'posthog'.
Audit Metadata