The Agent Skills Directory

[SAFE]: The skill reads ${CLAUDE_PLUGIN_ROOT}/.local.md to access a CV profile for context. This is expected behavior for a document assistant requiring user-specific data to perform its primary function.
[SAFE]: The skill processes data from $ARGUMENTS and user responses to AskUserQuestion. While these are external input surfaces, the skill lacks high-risk capabilities such as network access or shell command execution, mitigating the risk of injection or exfiltration.
[SAFE]: No obfuscation, remote code execution, or persistence mechanisms were detected in the skill instructions.

Document Assist