The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard, non-malicious shell commands (ls, find, du, mkdir, mv, md5) to perform filesystem analysis and organization. These operations are directly aligned with the skill's primary purpose.
[COMMAND_EXECUTION]: Strong safety protocols are embedded in the instructions, specifically requiring the agent to present a 'Proposed Structure' plan and obtain user confirmation ('Ready to proceed?') before executing any filesystem modifications or deletions.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external data in the form of filenames and file metadata across the user's filesystem.
Ingestion points: Target directory listings processed via ls, find, and file commands.
Boundary markers: The skill uses an approval-based workflow where the agent must summarize findings and propose a plan before taking action, creating a natural verification boundary.
Capability inventory: Filesystem write access (mkdir, mv) and deletion capabilities as part of the organization workflow.
Sanitization: Relies on human review of the generated 'Organization Plan' to validate that the agent's interpretation of the filesystem data is correct and safe.

file-organizer