Aida
Captions for emotional subtext — an accessibility tool that reads what people mean beneath their words, for neurodivergent people.
Core trick
Scores each outgoing message for emotional subtext against per-person history and shows a caption-style interpretation before sending.
Tags
Description
Captions for emotional subtext — an accessibility tool that reads what people mean beneath their words, for neurodivergent people.
README excerpt
# Aida — closed-captions for emotional subtext Aida is an **accessibility tool** for written conversation. It reads what people *mean* beneath what they *say*, and it catches a wounding message before you send it. The name (間, *aida*) is Japanese for "the space between" — the interval between two people, where meaning lives but words don't. Aida sits in that space and reads it aloud. If captions make sound legible and screen readers make pixels legible, Aida makes **emotional subtext** legible. That is the whole of it. **Live demo:** https://aida-yxi3.onrender.com — a working sample. (Free hosting, so the first request after idle takes a few seconds to wake; the in-memory store resets on each deploy, so add a person and a couple of their messages to see Aida learn them.) --- ## What it is - An assistive layer for **written** human-to-human conversation, in the lineage of captions and screen readers. - A reader of **subtext** — the emotion under the surface words — grounded in the specific person you are talking to. - A **safety gate** on what you send: it mirrors the emotion your own draft carries, and stops you before a message lands as a wound. - An **inline overlay** on the conversation. A quiet marker you tap to expand; a pulse only when it matters. ## What it is NOT - **Not mental-health or therapy.** No wellbeing advice, no diagnosis, no treatment. It translates communication; it does not counsel you. - **Not personality-typing.** It never types, scores, or lab