The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from user-provided reference images and project idea files to generate design systems and PRDs using a subagent. This content is eventually used to compose the final implementation prompt that drives code generation.
Ingestion points: User-provided reference images directory (Step 1/2) and project idea files (Step 1/3).
Boundary markers: No specific delimiters or "ignore instructions" warnings are used when passing external content to the subagent templates (assets/design-system.md, assets/app-overview-generator.md).
Capability inventory: The skill performs file system reads/writes, invokes subagents, executes shell commands (find/grep), and implements UI code in a React project.
Sanitization: No sanitization or validation of the input text or image metadata is performed before interpolation into subagent prompts.
[COMMAND_EXECUTION]: The skill executes shell commands to verify the project environment.
Evidence: Step 5 executes find . -name "package.json" -exec grep -l "react" {} \; to locate React projects.
[EXTERNAL_DOWNLOADS]: The skill suggests the installation of external dependencies if a React environment is not detected. These are well-known packages from official registries.
Evidence: Step 5 suggests running npm install -D tailwindcss postcss autoprefixer and npm install lucide-react.

ui-designer