exploring-llm-clusters
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses official PostHog tools (
posthog:execute-sql,posthog:query-llm-trace, etc.) to interact with observability data. All SQL queries are structured to access specific AI-related event tables and properties. - [SAFE]: The Python script
scripts/print_clusters.pyis a utility that parses JSON data from a file to print a human-readable summary of clusters. It does not perform network operations, access sensitive files, or execute arbitrary code. - [COMMAND_EXECUTION]: The skill provides SQL templates for querying metrics. These are standard analytical queries intended for use with the
posthog:execute-sqltool and do not show signs of SQL injection vulnerabilities or attempts to bypass database constraints. - [DATA_EXPOSURE]: The skill accesses
$ai_*event data, which is the intended purpose of an AI observability skill. It includes guidance on accessing model inputs/outputs viaposthog.ai_events, which is a documented part of the PostHog schema.
Audit Metadata