modeling-product-usage-metrics
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of analytical SQL templates and documentation. No malicious patterns such as credential exfiltration, obfuscation, or unauthorized command execution were detected.
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for processing event data, which represents a potential attack surface for indirect prompt injection.
- Ingestion points: The skill ingest event data from PostHog tables or dbt staging views (e.g.,
stg_events). - Boundary markers: The skill instructions (Rule 6 in SKILL.md) explicitly warn that event names are untrusted and should be treated as quoted data.
- Capability inventory: The skill facilitates SQL query execution and dbt model creation.
- Sanitization: The skill implements a security best practice by requiring user confirmation before creating persistent views based on event data, effectively mitigating the risk of automated injection.
Audit Metadata