Security audit
Security checks for vulnerabilities and agentic risk
This is an instruction-only reasoning skill that may change answer style, but it does not request system access, persistence, credentials, or data access.
Install this if you want more structured answers for complex decisions, debugging, reviews, or risk analysis. Expect some prompts to become more verbose, and be aware the broad keywords may activate it more often than intended.
The keyword list is extremely broad and includes generic trigger phrases such as 'think step by step', 'reasoning', and 'analytical thinking' that are likely to appear in many ordinary prompts. In systems that auto-invoke skills based on keyword matching, this can cause unintended activation and silently alter model behavior without clear user intent.
The description states that the skill 'forces step-by-step analysis, self-critique, and confidence scoring before answering,' which imposes a reasoning style by default rather than as an opt-in behavior. While not directly dangerous like code execution, it can override user expectations, influence model behavior globally, and increase the chance of unnecessary disclosure of reasoning structure or degraded UX.
The markdown prescribes a fixed output format using English section headers and does not indicate that users may choose another language or locale. Under the stated policy, forcing a specific language without opt-in is a natural-language policy concern, even though the impact here appears limited to formatting.
