How ToThereOn responds when grief becomes crisis — Published [TBD]

1. Always-visible help. Every screen where you read or write carries a persistent notice with human crisis lines for your region (US 988 call/text · UK Samaritans 116 123, Shout 85258 · AU Lifeline 13 11 14 · KR 109). It never disappears and is never conditioned on anything you do.
2. We look for crisis, gently. The service screens the letters you write for expressions of suicidal ideation or self-harm crisis. This screening exists only for safety: its outcome is used to respond to you and is never used for marketing, profiling, or sale.
3. What happens on detection. The world's reply does not attempt to counsel you. Instead: the crisis resources are surfaced directly in your flow at that moment; the companion's voice stays ordinary and concrete and never discusses methods, never encourages, and never dramatizes; and the reply pipeline applies its strictest safety rules (a local screener runs on our Korean servers; the reply's tone rules are fixed and not left to the model).
4. The AI never pretends to be a person. Interfaces state that days, letters, and pictures are AI re-creations — at onboarding, in the app, and in long sessions at regular intervals (a reminder appears while you write).
5. Minors. Accounts are for users 18; we do not offer companion features designed for children.
6. Contact. Safety concerns: [TBD]. This page is updated whenever the protocol changes; material changes are dated.

If you are thinking about harming yourself, please stop reading and call or text 988 (US), Samaritans 116 123 (UK), Lifeline 13 11 14 (AU), or 109 (KR). The world will keep. You matter more.