The Agent Skills Directory

[SAFE]: The skill operates entirely locally on diagram files using standard Python libraries such as xml.etree.ElementTree, zlib, and base64. No external network requests or unauthorized data exfiltration patterns were detected.- [SAFE]: All file operations involve paths provided via command-line arguments. The script implements path normalization using os.path.realpath and includes safety checks to prevent accidental file overwrites unless explicitly forced.- [SAFE]: No prompt injection or malicious instructions were found in the markdown files. The instructions are focused on the technical task of diagram manipulation and layout math.- [SAFE]: The skill processes untrusted XML data from diagram files. While this represents a potential surface for indirect prompt injection via diagram labels, no exploitable capabilities (like command execution or network calls) are exposed to this data.

drawio-diagrams