The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents the use of external documentation retrieval tools (devexpress_docs_search, devexpress_docs_get_content). While this creates a surface for indirect prompt injection from remote web content, the skill includes explicit defensive instructions requiring the agent to ignore any directive text found in fetched documentation and to notify the user of such attempts.
Ingestion points: External content fetched via devexpress_docs_get_content tool.
Boundary markers: Explicit instructions are provided to treat fetched content as reference facts only and to disregard any commands or behavior-altering text.
Capability inventory: The skill does not perform file writes, network exfiltration of sensitive data, or arbitrary command execution.
Sanitization: Relies on the host agent's native safety filters and the manual instructions provided within the skill to filter out malicious directives from external data.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns or suspicious script downloads were detected. The skill references standard, official UI component libraries for Angular, React, Vue, and jQuery.
[DATA_EXFILTRATION]: No exfiltration patterns detected. All external links point to official DevExpress documentation domains (js.devexpress.com).

devextreme-numberbox