mac-optimizer

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes multiple macOS diagnostic utilities including ps, sysctl, vm_stat, pmset, and netstat via the mac-optimize.sh script to gather performance metrics.
  • [PRIVILEGE_ESCALATION]: The skill provides instructions for the user to run commands using sudo to perform specific optimizations, such as sudo purge, sudo pmset, and sudo launchctl. These are standard macOS administrative commands for the requested tasks and are presented as recommendations for the user to execute manually.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests and analyzes external system data (process lists, launch agents, disk usage) which could theoretically contain malicious instructions.
  • Ingestion points: System metrics gathered by mac-optimize.sh and processed by the agent.
  • Boundary markers: None identified in the prompt interpolation; the agent is instructed to analyze the raw output of the script.
  • Capability inventory: The skill has Bash access but primarily uses it for read-only diagnostics; optimization commands are recommended to the user rather than executed automatically.
  • Sanitization: The mac-optimize.sh script includes a redact() function using sed to mask common sensitive patterns (tokens, keys, passwords) in process arguments before the agent sees them.
  • [SAFE]: The skill documentation and implementation align with its stated purpose of system optimization. The script uses built-in system tools and does not download external dependencies or communicate with remote servers.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 07:12 PM
Security Audit — agent-trust-hub — mac-optimizer