The Agent Skills Directory

[SAFE]: The skill is entirely documentation-based, providing patterns and rules for developers. It does not contain any executable code, scripts, or instructions that attempt to bypass safety filters or exfiltrate data.
[SAFE]: The instructions explicitly enforce multi-tenancy security by requiring that all backend queries include a businessId filter, which prevents data leakage between different business entities.
[SAFE]: Access control is well-defined through a permission matrix (purchases.read, purchases.write, and admin), ensuring that sensitive operations like deletion are restricted to authorized users.
[SAFE]: Input validation rules are documented for purchase creation and status updates, including existence checks and range validations for quantities and costs.

purchases-flow