What Reprogram Within will and will not say

For everyone

• Every line the app plays is positive, present tense and calm. It refuses negations such as "I am not a failure", because rehearsing the negative keeps it in the room.
• It will not voice anything scary, violent, sad, tense, shaming or self-critical, and will not voice diagnosis, treatment or medical claims.
• A request containing self-criticism is never repeated back. The app offers a kinder reframe instead.
• A request that suggests acute distress is never answered with a script. The app shows support lines for your region.
• Nothing is inaudible. The whisper layer is your own recorded words, readable on screen, with a "hear it loud" button.

For children

• Only pre-written scripts from the vetted library, or a parent's own words that pass the same checks, are ever played.
• Both-parent-safe. Nothing played to a child may speak badly of a parent, step-parent, grandparent, sibling or home, favour one home over another, or suggest only one person loves them. This holds for the library, the composer and anything a parent types.
• No secrets from parents, in any direction. No romantic or adult themes, no mention of diets, weight or substances.
• Every recording from someone other than the parent is held until the parent approves it.
• The library is reviewed by a child psychologist before public release.

Voices

• Family voices are recordings made on the device and kept there.
• If voice cloning is ever offered, it will be for consenting adults only, with recorded consent and a watermark. The app will never clone a child's voice.

Custom words

• Your own words go through the same checks as composed text before they can be recorded. A refused line is shown with the reason and a kinder alternative.
• The app logs the decision as a one-way hash of the text, never the text itself.

If something gets through

Tell us: [support@reprogramwithin.com]. We treat it as a bug, fix the rule, and add the case to the tests.