creating-online-evaluations

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFE

Full Analysis

[COMMAND_EXECUTION]: The skill uses posthog:execute-sql to perform database queries for the purpose of verifying event volumes and testing property filters.
[COMMAND_EXECUTION]: The skill facilitates the deployment of Hog source code via posthog:llma-evaluation-create to automate the scoring of events.
[CREDENTIALS_UNSAFE]: The skill retrieves and configures LLM provider key identifiers using posthog:llma-provider-key-list to enable LLM-based evaluations.

Audit Metadata

Risk Level

SAFE

Analyzed

Jul 24, 2026, 02:29 PM

Security Audit — agent-trust-hub — creating-online-evaluations