The Agent Skills Directory

[SAFE]: The skill is entirely instructional and provides configuration templates for a database-backed platform. It does not contain any executable code, external network requests, or attempts to access sensitive local system data.
[PRIVILEGE_ESCALATION]: The documentation references the use of SECURITY DEFINER for PostgreSQL triggers. This is a standard database design pattern used to allow users to trigger logic (like earning an achievement) that writes to restricted tables (like credits) without granting the user direct write access to those tables. This follows the principle of least privilege within the database context and is presented as part of the intended architecture by the vendor.
[INDIRECT_PROMPT_INJECTION]: While the skill defines a system that processes data (row changes) to trigger events, it provides a structured schema for conditions and requirements. As a design for a backend module within the vendor's ecosystem, it does not expose the agent or the host system to external injection attacks. Findings in this category are assessed as safe as they represent the primary intended purpose of the skill.

constructive-events