The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to ensure the local development environment is ready for testing.
Evidence: Uses Test-NetConnection (on Windows) or lsof (on POSIX) to check if the local server on port 3000 is active.
Evidence: Invokes pnpm dev with isBackground: true to start the development server if the port is not responsive.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests and processes content from local web pages during validation.
Ingestion points: Data enters the agent's context through browser_navigate, take_snapshot (accessibility tree), and evaluate_script (CSS styles).
Boundary markers: No specific delimiters or warnings are used to distinguish between system instructions and content retrieved from the web page.
Capability inventory: The agent has the ability to execute shell commands and perform arbitrary browser interactions.
Sanitization: No explicit sanitization or filtering of the page content is performed before the agent analyzes it for UI compliance.

ui-validator