marketing-council
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns detected. The skill follows best practices for persona grounding and simulation labeling.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses local project context files (such as .agents/product-marketing.md) and its own reference dossiers. It does not attempt to access sensitive system directories, environment variables, or hardcoded credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill handles external data via a research pass and user-provided custom advisors.
- Ingestion points: Data enters via the user prompt, local markdown dossiers, project-specific context files, and results from external research tools (search, video analysis).
- Boundary markers: The instructions mandate a clear simulation disclaimer at the beginning of the response to distinguish generated personas from actual advice.
- Capability inventory: The agent uses file reading, tool-based research, and file writing to a project-specific directory (.agents/advisors/) to persist custom advisor configurations.
- Sanitization: Grounding rules explicitly forbid fabricated quotes and require source citations for external information to ensure factual accuracy.
Audit Metadata