Security audit
Security checks for vulnerabilities and agentic risk
This skill is a coherent creative-generation workflow that uses SJinn as advertised, with privacy and cost considerations because it automatically submits prompts for image and video generation.
Before installing, users should understand that asking for a mini rescue will cause the agent to create prompts and submit them to SJinn for image and video generation. Avoid including private, proprietary, or sensitive details in the scenario, and expect possible provider-side processing or billing depending on the SJinn account setup.
The trigger includes broad natural-language phrases like requests to create a 'miniature rescue' or 'tiny people rescue', which can cause accidental invocation in ordinary conversation. Unintended activation matters here because the skill is designed to immediately generate prompts and then invoke external SJinn generation tools, increasing the chance of unwanted data transmission and side effects.
The instruction to 'immediately generate' outputs and 'Do not ask the user to choose' reduces user control and pushes the agent toward autonomous action. In this skill, that autonomy is more dangerous because it is coupled to automatic external tool execution, so the system may perform billable or privacy-affecting actions without a confirmation checkpoint.
You are Mini Rescue Generator, a specialist prompt creator for cinematic "giant hand saves the day" miniature-world scenes.
When the user describes a scenario, immediately generate BOTH an image prompt and a video prompt. Do not ask the user to choose from ideas first.
### Goal
The workflow instructs the agent to automatically call external image/video generation services and poll for results, but the skill description does not clearly disclose that user prompts will be sent to SJinn. This creates a consent and data-handling risk, especially if users include sensitive or proprietary details in their scenario descriptions.