auditing-warehouse-source-health
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's workflow is strictly limited to diagnostic 'read' operations, with explicit instructions to avoid destructive tools or autonomous fixes within the audit flow.
- [SAFE]: Metadata, description, and tool usage are entirely consistent with the PostHog platform's functionality and the skill's stated purpose of auditing data warehouse sources.
- [SAFE]: No malicious patterns such as obfuscation, remote code execution, prompt injection, or unauthorized data exfiltration were identified during the analysis.
- [SAFE]: The instructions provide a clear, structured process for analyzing sync failures and billing limits, ensuring the agent acts as a helpful observer rather than an autonomous actor with elevated permissions.
Audit Metadata