skills/boshu2/agentops/sbh/Gen Agent Trust Hub

sbh

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill interacts with the host system using the sbh command-line utility. It executes commands to inspect mount points, perform status checks, and run authorized cleanup or configuration tasks.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from the storage environment to make recovery decisions, representing an indirect injection surface.
  • Ingestion points: Data from sbh --json status and sbh check output (SKILL.md).
  • Boundary markers: The skill instructions explicitly require obtaining caller authority and establishing a factual baseline before any mutation (SKILL.md).
  • Capability inventory: The skill possesses capabilities for file deletion, ballast release, and storage configuration modification (SKILL.md).
  • Sanitization: The skill mandates an evidence-first approach, prioritizing reversible actions and strictly adhering to file protection vetoes (SKILL.md).
  • [EXTERNAL_DOWNLOADS]: The skill includes a reference to a GitHub repository (github.com/Dicklesworthstone/storage_ballast_helper) for documentation and tool source information.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 07:34 AM
Security Audit — agent-trust-hub — sbh