OverviewHistoryStatsSecurity
Security
Independent checks from skills.sh's audit partners.
Independent checks from skills.sh's audit partners.
Independent checks from skills.sh's audit partners.
Warn3 providerslatest audit Sep 10, 2026
Independent checks from skills.sh's audit partners.
This skill provides templates for building AI agents with the ability to control a computer's mouse, keyboard, and shell. While it includes robust security advice such as sandboxing, resource limits, and human-in-the-loop confirmations, the core functionality introduces a high-risk attack surface, specifically through shell command execution and potential exposure to indirect prompt injection from untrusted screen content.
Detected behaviors
1 alert: gptSecurity
1 issue