The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches a component registry manifest from the official vendor domain at https://ui.inference.sh/r/widgets.json to configure the UI blocks.\n- [COMMAND_EXECUTION]: The documentation provides instructions to execute shell commands using npx to install the skills utility, the belt-sh/cli package, and the shadcn component registry.\n- [INDIRECT_PROMPT_INJECTION]: The WidgetRenderer component generates interactive UIs (including forms, inputs, and buttons) from structured JSON responses. This creates a vulnerability surface where malicious instructions or deceptive elements could be injected if the agent processes untrusted external data.\n
Ingestion points: The widget data object passed to the WidgetRenderer component as documented in SKILL.md.\n
Boundary markers: No explicit delimiters or validation warnings are provided for the input data in the documentation.\n
Capability inventory: Full interactive UI rendering, including forms with submit actions and buttons with click handlers.\n
Sanitization: The documentation does not describe any specific sanitization or schema validation processes for the input JSON before it is rendered to the user interface.

widgets-ui