Security audit
Security checks for vulnerabilities and agentic risk
This is a markdown-only writing skill for self-help content, with no code execution or data access, but users should verify its psychology and health-related claims.
Install only if you want a strong self-help writing style. Review outputs for accuracy before publishing, verify cited studies and biographical stories, and do not rely on generated content as medical, psychological, psychiatric, or crisis advice.
The trigger instructions are unusually broad and include catch-all language like 'always use this skill' and 'even casual mentions,' which can cause the skill to activate for loosely related requests. That increases the chance the system applies strong writing constraints when the user did not explicitly ask for them, leading to misrouting, unwanted behavior, or suppression of better-matched skills.
The skill imposes a default second-person voice convention without explicit user opt-in, which can override user-preferred style or system-level orchestration choices. In isolation this is not a security-critical flaw, but it is still a policy and control issue because hard stylistic mandates reduce user agency and can cause inappropriate outputs in contexts where a different voice is required.
This reference file presents psychology, neuroscience, stress, loneliness, and wellbeing claims in a way that can be reused as authoritative self-help guidance, but it does not clearly warn that the material is informational and not medical, psychological, or mental-health advice. In the context of a skill designed to generate polished self-help content, users may over-trust these statements and apply them to health or mental-health decisions, especially where the text uses strong causal or biological framing such as 'it is medicine' or 'change is biological.'
No suspicious patterns detected.
