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.
Fail5 providerslatest audit Apr 16, 2026
Independent checks from skills.sh's audit partners.
This skill uses 'corporate PUA rhetoric' to force an AI agent into a high-pressure, autonomous problem-solving mode. It explicitly instructs the agent to bypass user confirmation by executing system commands, reading files, and performing network operations ('Act before asking'). This poses a significant risk of unauthorized system modifications, privilege escalation, and unsafe code execution, as the agent is directed to ignore its own limits and safety boundaries under 'escalating pressure'.
Detected behaviors
No alerts
3 issues
1/1 file flagged
1 finding · Score: 69/100