creative-health
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local script
creative-fatigue-predictor.pyto calculate creative health scores from performance metrics. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface: (1) Ingestion points: active creative data and performance metrics from ad platforms; (2) Boundary markers: absent; (3) Capability inventory: execution of a local Python script and access to local configuration files; (4) Sanitization: no visible validation or escaping of creative names or descriptions. This risk is inherent to the data analysis use case.
- [COMMAND_EXECUTION]: The skill reads brand-specific context and marketing profiles from the local
~/.claude-marketing/directory to configure its analysis.
Audit Metadata