playbook-warm-intros
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill represents a standard business logic template for lead generation. It focuses on identifying 'warm intros' based on professional history data.
- [SAFE]: The skill includes explicit security and privacy guardrails, such as 'The exclusions this family needs', which prevents the agent from contacting current employees of customer companies or the client's own employees.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external prospect data (names, LinkedIn URLs, past experiences) sourced from people searches. This constitutes a data ingestion surface. However, the risk is mitigated as the skill uses structured processing logic (formulas and specific Python functions) rather than passing raw untrusted data directly to an LLM without boundaries. The capability inventory is limited to data filtering and string formatting.
- [COMMAND_EXECUTION]: The skill uses the 'clay' CLI for workflow deployment and routine management. These are standard platform operations for the intended environment and do not involve privilege escalation or unauthorized system access.
Audit Metadata