Safety and scope
Small, deliberate limits.
This is a parent-directed educational worksheet preview for ages 5–15. Its checks reduce some risks, but they are not a dedicated child-safety system or a promise of complete safety.
Prompt handling
Scene descriptions are checked for several categories, including sexual or nude material, graphic violence, self-harm, dangerous or drug-related requests, hateful content, real-person likenesses, personal details, prompt-injection attempts, and named characters. Rejected requests receive an error; the app does not silently substitute a different scene.
These checks use fixed text patterns rather than a general-purpose moderation model. They can miss context and may reject benign wording. Do not include a child's name, age, school, location, contact details, or other personal information.
Artwork handling
Art is composed from local SVG drawing rules; no image-generation provider is called. The free-form text steers a limited set of recognizable subjects and scenery, and may not be represented exactly. A structural output check runs before the preview or download is returned.
The learning activity uses deterministic authored content with answer checks, but it is not a substitute for a teacher or a curriculum review. An adult should review the art and exercise before sharing or printing.
What V1 does not support
- Child accounts, independent child use, chat, or direct messages.
- Photo uploads, face likenesses, or transformation of a child's image.
- Public prompts, galleries, social sharing, or family sharing.
- Open-ended image generation, guaranteed moderation, or guaranteed educational outcomes.
- Facebook Login, Meta Pixel, Meta Conversions API, or Meta advertising.
Qualified privacy counsel should review the actual audience, data flows, and provider terms before any independent child use or collection of user information is permitted.