disk-manager
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local shell script
{baseDir}/disk.shto perform disk analysis, search for files, and manage storage. - [COMMAND_EXECUTION]: Includes a
--deleteflag that allows the agent to perform destructive filesystem operations, such as removing old or temporary files. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data (filenames and directory paths) from the host system that could potentially contain malicious instructions.
- Ingestion points: Filesystem metadata is read from the local system through commands like
du,find, anddf. - Boundary markers: No delimiters or specific instructions to ignore embedded content within filenames are provided.
- Capability inventory: The skill can execute shell commands and delete files via the
disk.shscript. - Sanitization: No input validation or sanitization for filenames or paths is documented for the underlying script.
Audit Metadata