Security audit
Security checks for vulnerabilities and agentic risk
This skill only changes response style for terse communication and does not request tools, files, credentials, network access, or system privileges.
Install only if you want terse replies and are comfortable disabling it with "stop caveman" or "normal mode" if it activates after a broad brevity request.
The skill advertises wenyan/classical Chinese output modes as part of normal operation, but the auto-trigger design does not require explicit user selection of those language variants. This can cause unexpected language switching or non-user-preferred output, reducing comprehensibility and increasing risk when handling instructions, warnings, or technical guidance.
The trigger set is broad enough to activate on ordinary requests like 'be brief' or 'less tokens', which can unintentionally force a persistent response-style change across later turns. Because the mode persists until explicitly disabled, accidental activation can degrade clarity, alter safety-critical messaging, and create confusing behavior without clear user consent.
No suspicious patterns detected.
