The Agent Skills Directory

[SAFE]: Detailed inspection of the skill instructions and accompanying scripts confirms that the code performs only the described XML parsing operations. No indicators of data exfiltration, obfuscation, or unauthorized access were detected.
[COMMAND_EXECUTION]: The skill executes local scripts (PowerShell or Python) to process configuration data. These operations are limited to the provided file paths and are standard for developer-oriented auditing tasks.
[PROMPT_INJECTION]: The skill handles external data inputs by parsing XML files. While this represents a surface for indirect prompt injection, the analysis is restricted to structural metadata extraction, and the results are presented as an informational audit summary.

role-info