The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The skill utilizes standard unix utilities (ls, find, du, mkdir, mv) to perform file management.
Evidence: Commands like find [directory] -type f -exec md5 {} \; are used locally to identify duplicates.
Context: These operations are directly aligned with the stated purpose of organizing files and folders.
[DATA_EXFILTRATION] (SAFE): While the skill reads file names and metadata, there are no network-capable commands (e.g., curl, wget, fetch) present to transmit data to external servers.
[PROMPT_INJECTION] (SAFE): No instructions attempting to override agent behavior, bypass safety filters, or extract system prompts were found.
[INDIRECT_PROMPT_INJECTION] (LOW):
Ingestion points: Reads file names and types using ls and find in SKILL.md (Step 2).
Boundary markers: None explicitly defined for file names, but instructions specify asking for confirmation before major actions (Step 4 & 6).
Capability inventory: File system write/delete via mv and rm (implied).
Sanitization: None detected for file names, posing a minor risk if a file is named with injection patterns, though the impact is limited by the interactive nature of the skill.

file-organizer