The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The documentation includes patterns where external data (API responses or error messages) is rendered into the DOM using innerHTML.
Ingestion points: Data from fetch() calls in SKILL.md (Pattern 1) and error messages in references/advanced-usage.md (Error Handling section) are rendered directly into container elements.
Boundary markers: No explicit boundary markers or security warnings are used around these interpolation points.
Capability inventory: The skill performs DOM manipulation via innerHTML and uses the fetch API for data retrieval; it does not have access to sensitive system files or subprocess execution.
Sanitization: The provided examples do not demonstrate sanitization or the use of safer alternatives like textContent for dynamic data rendering.
[UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill documentation references external resources for installation and setup.
Evidence: Recommends cloning the quickstart repository from github.com/SyncfusionExamples/ej2-quickstart-webpack and fetching styles from cdn.syncfusion.com.
Context: These resources belong to the official vendor and are documented for standard development workflows.

syncfusion-javascript-spinner