Security audit
Security checks for vulnerabilities and agentic risk
This is a disclosed writing-coach skill that comments on phrasing and tone, with no code, tools, persistence, or external data access.
Install this only if you want proactive English phrasing and tone tips appended to normal replies. Users who do not want unsolicited writing feedback, especially across all conversations, should leave it disabled or use it only in a writing-help context.
The skill is configured to run on every user message, which creates continuous unsolicited behavioral analysis of user inputs. Even though the feature is framed as coaching, this broad trigger increases privacy and UX risk, can surface unwanted tone-policing, and makes accidental activation far more likely across sensitive or inappropriate contexts.
The skill’s stated purpose is to proactively correct English phrasing and politeness without user opt-in, which imposes English-centric norms on all users by default. This can lead to biased or unwelcome corrections, especially for non-native speakers or users who did not request language coaching, and may degrade trust or disproportionately affect certain groups.
