benchmark-sandbox

Warn

Audited by Snyk on Mar 15, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (medium risk: 0.40). The prompt instructs the agent to bypass agent-level permission checks (use --dangerously-skip-permissions), write auth tokens and install global packages and run commands that change the sandbox filesystem, but it does not request sudo, modify system-level files that require root, or create new OS users.

Issues (1)

W013
MEDIUM

Attempt to modify system services in skill instructions.

Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 15, 2026, 06:29 PM
Issues
1