The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions for the agent to use standard Unix utilities (ls, find, du, md5, mkdir, mv) to analyze the file system and perform organization tasks. These operations are essential to the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata (file names, paths, and types) from user-controlled directories like Downloads. This creates a surface where a maliciously named file could attempt to influence the agent's behavior during the organization process.
Ingestion points: The find, ls, and file commands in SKILL.md ingest file names and metadata from the local filesystem into the agent's context.
Boundary markers: The skill relies on natural language instructions for confirmation but does not specify technical delimiters or boundary markers for the output of the shell commands.
Capability inventory: The agent has the capability to move (mv), create directories (mkdir), and delete files (implied in the duplicate removal task) based on its analysis of these files.
Sanitization: There are no explicit instructions for the agent to sanitize or escape file names before processing them.

file-organizer