The Agent Skills Directory

[COMMAND_EXECUTION]: The 'export json' command in 'scripts/joplin.sh' executes Python code via 'python3 -c' using shell interpolation for the '$NOTES_DIR' variable. This pattern allows for dynamic code execution if the environment variable is manipulated to contain Python string delimiters (e.g., single quotes).
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by reading user-controllable data from local note files. Ingestion points: Note content is read and displayed to the agent in the 'view', 'search', and 'list' commands. Boundary markers: No delimiters or explicit warnings are provided to the agent to disregard instructions within note content. Capability inventory: The script uses shell commands and local Python execution to manage files and data. Sanitization: Content from files is read and presented directly without sanitization or escaping.

note-organizer