analyzing-expensive-users

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses posthog:execute-sql to perform complex analytical queries on event data. These queries are well-structured for the intended purpose of AI cost observability and include built-in filters for time ranges and specific event types ($ai_generation).
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill contains specific instructions to protect personal data, explicitly warning the agent: 'Never select the raw person.properties object... it serializes the full property blob into the result and leaks personal data far beyond a label.' It uses argMax to select only specific identifying labels (email, name) instead of broad property access.
  • [SAFE]: All tools (posthog:*) and data sources (internal PostHog events) are consistent with the skill's stated purpose of providing AI observability for PostHog users. The guidance provided for optimization (caching, model mix, context reduction) is standard industry practice for cost management.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 09:39 AM
Security Audit — agent-trust-hub — analyzing-expensive-users