The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill utilizes aws and docker commands to build, push, and deploy container images. These operations are core to the skill's stated purpose and are performed with user confirmation.
[PROMPT_INJECTION] (LOW): The skill is susceptible to indirect prompt injection (Category 8) because it extracts values like service names and environment variables directly from the user's local codebase for use in shell commands. Ingestion points: User's local directory name and application source code. Boundary markers: Absent. Capability inventory: docker build, docker push, and aws ecs management commands. Sanitization: No explicit sanitization of extracted codebase strings before shell interpolation is defined.

ecs-express