file-organizer
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of filenames, file types, and metadata from the user's local directories. A maliciously crafted filename could attempt to influence the agent's logic during the analysis or proposal phases.
- Ingestion points: Directory contents and file metadata are read using
ls,find, andmd5commands as specified inSKILL.md. - Boundary markers: The skill includes instructions to present a clear organization plan and obtain explicit user confirmation before executing any changes.
- Capability inventory: The agent is instructed to perform file system operations including
mkdir,mv, and file deletion for duplicates. - Sanitization: The provided shell command examples utilize double quotes for file paths to help prevent simple command injection from filenames.
- [COMMAND_EXECUTION]: The skill relies on the execution of shell utilities to perform its primary tasks. The agent dynamically constructs and executes these commands based on the state of the user's file system and the target directories provided.
Audit Metadata