The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill uses the !command`` syntax in SKILL.md to load external protocol files and project-specific configurations into the agent context (e.g., cat .production-grade.yaml). These operations are benign, as they are used to establish ground rules and context from local files within the project directory.
[DYNAMIC_EXECUTION]: The provided code templates for backend services and analytics pipelines follow industry-standard security practices, including the use of security headers, rate limiting, and structured audit logs. No unsafe execution of untrusted data was found.

liveops-engineer