hundred-million-leads
Pass
Audited by Gen Agent Trust Hub on Mar 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses a structured multi-phase workflow to guide users through a business workshop. The instructions are focused on strategy and role-play without any attempts to bypass agent safety filters or override system constraints.
- [SAFE]: Web search operations are explicitly scoped to market research, such as analyzing competitor tactics, platform demographics, and industry benchmarks. No patterns of sensitive data harvesting or unauthorized network exfiltration were found.
- [SAFE]: The use of subagents ('general-purpose' task agents) is a standard orchestration technique for providing diverse feedback (e.g., Skeptical Marketer, Business Strategist). These agents operate within the platform's simulated environment and do not execute external code or scripts.
- [SAFE]: The 'Customer Personas' feature is a simulation where the AI role-plays as a potential client to review outreach scripts. This is a legitimate application of LLM capabilities and does not involve real-world contact with third parties or data leakage.
Audit Metadata