The Agent Skills Directory

[SAFE]: The skill is primarily composed of educational content, process guides, and architectural blueprints for CRM management. No indicators of prompt injection, data exfiltration, or persistence were found.
[COMMAND_EXECUTION]: The skill includes a benign Python script (scripts/check-output.py) used locally to validate that generated deliverables contain required markdown headers. The script performs simple string matching and contains no dangerous network or system operations.
[DATA_EXFILTRATION]: No network operations or unauthorized data access patterns were identified. While the skill mentions integration with enrichment services like LeadMagic and Clay, these are described as standard business workflows for the end-user's CRM environment, not as actions for the agent to perform to exfiltrate data.

crm-toolkit