The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to perform standard development environment operations such as git init and creating project directories and files. These actions are aligned with the skill's primary purpose of project initialization.
[INDIRECT_PROMPT_INJECTION]: The skill contains logic to ingest and process user-provided data, including existing prototype files, design specifications, and reference URLs. While this presents an attack surface for indirect prompt injection, the skill mitigates this by requiring explicit extraction steps, human review checkpoints, and verification against defined design contracts.
[DYNAMIC_EXECUTION]: The skill generates high-fidelity interactive prototypes in the form of single-file HTML/CSS/JS artifacts (demo.html). These are intended for user verification in a browser and are not executed in a privileged environment or used for server-side code generation within the shaping phase.
[SAFE]: The skill demonstrates security-conscious behavior by recommending the use of .env.example for secret management and emphasizing that it should not write or replace business logic or backends with mocks, thereby maintaining a clear boundary between design prototyping and application code.

fancy-shape