Start here
The free tool below runs entirely in this browser — no account, nothing stored or transmitted. Redact anything you would not want kept before you type it.
Open the companion tool →Why fluency is not the same as fit
A general chatbot answers from patterns in language. It has never met your child, cannot see a rash, does not know your family history, and cannot notice that you are exhausted. For general information that is fine; for anything specific to one child it is a limit, not a flaw.
The questions that go wrong most often are the ones that feel small: a sleep change, a new behavior, a phrase that worried you once. Those are exactly the questions where a pediatrician, a teacher, or a counselor who knows the child has information the model does not.
The protocol
- Write the question the way you would ask a person, including the details that matter: age, how long, what changed, what you already tried.
- Ask what kind of answer this needs. General information about child development is a reasonable chatbot question; anything about your specific child's health, safety, or behavior is not.
- If a child is in immediate danger or needs medical care, skip the tool entirely and contact emergency services or a clinician.
- For everything else, write your one-sentence question down and bring it to the person who sees your child: a pediatrician, a teacher, a school counselor, or a family member who knows them.
- Keep the chatbot's answer, if you keep it, as background reading rather than as the decision.
What the tool does
Describe the question you were about to ask and get a boundary reading: what a general assistant can help with, what it cannot see, and who the question belongs to. It runs in this browser; nothing is sent or stored.
Limit first: This is a boundary-reading aid, not medical, developmental, or mental-health advice, and it cannot assess your child. A clinician or a professional who knows your child is the right source for anything specific.
Open the companion tool →One human next step
Bring the one-sentence question to a pediatrician, a teacher, or a counselor at your next opportunity — and sooner if something changed suddenly. If a child is in immediate danger, call emergency services.
Research log and safety checks
AI-assist path: not used for a verdict. The reading is a boundary sort between general information and person-specific judgment; the whole value is that the reader sees the boundary, not a model's opinion of their parenting question. §6.10: not a frontier-pace topic.
§0.5 protection result: category is chatbot-dependency-check; crisis-adjacent: no; fear/urgency/scarcity toolkit used: no. The protective function is free and the paid feature is convenience only.