sb-hub

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected. The skill follows best practices for state management and local environment diagnostics.
  • [DATA_EXPOSURE]: The issue reporting utility (report-issue.sh) explicitly sanitizes data before drafting a report. It omits string values, source code, and token values, capturing only numeric counts and structural keys to protect project privacy.
  • [COMMAND_EXECUTION]: Shell and Python operations are used for legitimate project state detection (probing for React, Vite, and Storybook configuration). Python logic is executed via hardcoded heredocs with shell expansion disabled, preventing injection risks.
  • [EXTERNAL_DOWNLOADS]: The skill references standard package managers (npm, pnpm, yarn) and well-known tools like npx storybook. No downloads from untrusted or unknown remote sources were found.
  • [PROMPT_INJECTION]: The instructions focus on routing and state detection without attempting to override agent safety protocols or system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 06:56 PM
Security Audit — agent-trust-hub — sb-hub