storage-audit
Warn
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPERSISTENCEPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill requires the agent to execute multiple local shell and Python scripts (e.g., './scripts/storage-hygiene.sh', 'scripts/disk-report.py') to automate system maintenance and reporting tasks.
- [PERSISTENCE]: The skill documentation describes and executes the installation of persistent background processes, including a macOS LaunchAgent ('com.matias.storage-hygiene') via 'scripts/install-storage-hygiene-agent.sh' and a scheduled cron job ('storage-hygiene-every-3-hours').
- [PRIVILEGE_ESCALATION]: The instructions involve interacting with system configurations like '/etc/fstab' and using 'launchd' for agent installation. The skill also performs process termination ('reaping') of background services using the 'kill' command.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted system data which could potentially contain malicious instructions or influence agent behavior.
- Ingestion points: System logs in '~/.local/state/matias/storage-hygiene/' and diagnostic reports in '/Library/Logs/DiagnosticReports'.
- Boundary markers: Absent; the instructions lack explicit delimiters to separate system data from agent instructions.
- Capability inventory: The skill has capabilities for file system modification, script execution, and process termination across multiple provided scripts (e.g., 'scripts/storage-hygiene.sh', 'scripts/disk-report.py').
- Sanitization: Absent; no sanitization or validation of the ingested system data is described in the prompt workflow.
Audit Metadata