Security audit
Security checks for vulnerabilities and agentic risk
This is a text-only image prompt writing skill with no hidden execution, persistence, credential use, or data access.
This skill appears safe to install for image-prompt assistance. Be aware that it will default to Midjourney and English prompt text unless you ask otherwise, and its built-in style examples may influence outputs unless you specify a different aesthetic.
The skill description is broad enough to activate on many generic creative or image-related requests, which can cause incorrect routing and unintended use of this skill outside its intended scope. In an agent system, over-broad dispatch increases the chance of prompt interception, policy bypass by misclassification, or degraded handling by stealing requests from more appropriate specialized skills.
The instruction requires the final prompt text to be in English unless the user explicitly requests another language. This imposes a language choice by default rather than offering the user a neutral language option or opt-in, which fits the locale/language policy violation criteria.
The default invocation phrase is very generic and closely matches ordinary user requests about creating or refining image prompts. This can cause the skill to trigger unexpectedly on routine conversation, expanding its scope without clear user intent and increasing the chance of prompt-routing misuse or accidental delegation.
The phrase "soft diffuse North-European daylight" recommends a specific regional locale aesthetic as a default prompt ingredient. Because the guide does not frame this as optional user preference or offer alternatives, it can be read as imposing a locale-specific style without opt-in.
The sample prompt uses "North-European daylight," which hard-codes a regional/locale-specific aesthetic in natural language. The file does not indicate that this should only be used when the user wants that style, so it may violate language/locale choice expectations.
No suspicious patterns detected.
