The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a bash script scripts/drawio-to-png.sh that executes shell commands to call external diagram conversion tools like drawio or drawio-batch. While the script uses quoting for variables, it is a mechanism for executing shell commands based on provided file paths.
[PROMPT_INJECTION]: The skill is designed to ingest and manipulate external .drawio XML files, which constitutes a surface for indirect prompt injection.
Ingestion points: Reads and edits .drawio XML files (e.g., in SKILL.md examples).
Boundary markers: No explicit instructions are provided to the agent to treat XML content as untrusted data or to use delimiters to prevent instruction leakage.
Capability inventory: The skill can execute shell scripts (drawio-to-png.sh), Python scripts (find-arch-icon.py), and perform XML modifications.
Sanitization: There is no evidence of sanitization or structural validation performed on the XML content before the agent processes it.

drawio-designer