How we keep quests safe

How a quest is made

1. A starter library. Each daily set draws on a hand-written library of quest ideas (puzzles, reading, math, making, movement and more), which the AI adapts to your child.
2. Your child's profile. Age, interests, school grade and your settings decide the tone and difficulty. A seven-year-old and a fifteen-year-old get different quests.
3. Written rules for the AI. The instructions the model follows say: nothing dangerous, scary, violent or adult; physical quests stay light; kitchen and experiment quests are safe and assume an adult is around; no quest asks a child to buy anything, install anything, leave the house unsupervised or use materials they are unlikely to have.
4. Format checks before anything is saved. Every answer from the model is checked against a strict format (subject, type, quiz questions and answers, and so on). Output that doesn't fit is retried and never shown to your child.
5. Rule checks with a repair step. The set is then checked against the day's plan, including the subjects you asked us to avoid. If it breaks a rule, the model is asked to fix exactly that before the quests are saved.

Separately, every time we change how quests are written, we run test generations for a range of sample child profiles through an automated quality check. Among other things it flags age-unsafe words and ideas — knives, stoves and ovens, lighters and matches, bleach, going out alone, installing apps, social-media accounts, contacting strangers online, sharing personal details, buying things online — so a change that makes quests riskier doesn't ship. The same word check also runs live on every quest before it is saved: a quest it flags is sent back once for a fix and dropped if it still fails, links to sites outside a short list are removed, and physical, cooking and outdoor quests for children under 12 always carry a line asking for an adult nearby. No filter is perfect, so your own glance at the day's quests still matters.

What you control as a parent

• Subjects to avoid or emphasize. Set any subject to “avoid” and the AI is told never to use it; a set that slips is sent back for repair.
• Feedback that sticks. Thumbs up or down, with an optional comment, on any quest. Feedback is summarized into notes that shape future quests.
• Your own quests. Add quests you write yourself — once, daily or weekly — alongside the generated ones.
• Your child's comments. You can see the comments your child leaves on quests in the parent area.
• Skip anything. No quest is compulsory, and skipping one costs nothing.

Supervision by age

AI-written quests are suggestions, and you know your child, your home and your street better than any model does. A rough guide, not a rulebook:

Photos stay in the family

Adding a photo when a quest is done is always optional. Photos are stored on our own server, never sent to the AI, and only served to the child who added them and the parents of that family, after a sign-in check. They are never public and never used for advertising.

No ads, no trackers, no strangers

• No advertising, and no third-party analytics or tracking pixels, on the app or this site.
• No chat with people outside your family, no public profiles and no feed.
• Children don't sign up or give us an email address; their profiles live inside the parent's account.

The privacy page lists exactly what is stored and what is sent to our AI provider, and the terms cover the rest.

Report a problem

If a quest looks unsafe, wrong or unsuitable, skip it, give it a thumbs down with a short comment, and email us at hello@tandomo.com with the quest title and your child's age (no other details needed). A person reads every report, and we use reports to tighten the rules and checks above.