The Agent Skills Directory

[SAFE]: The skill provides structural instructions for technical documentation and planning. It defines a rigid format for the AI to present an 'implementation-ready sketch' using ordered phases and specific bullet hierarchies.
[SAFE]: It includes explicit constraints (e.g., 'Do not modify application code, tests, configuration, or project assets') that prevent the agent from performing potentially harmful file modifications during the session.
[SAFE]: No suspicious patterns such as obfuscation, network exfiltration, remote code execution, or credential harvesting were found. The examples provided use standard Elixir/Phoenix project structures as benign templates.

sketch-solution