# toba_apex

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Toba Falade](https://cerebralvalley.ai/u/toba_apex)
- **GitHub:** https://github.com/tobacreator/somamap
- **Demo video:** https://youtu.be/NC8UF-6Nzok
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4295

SomaMap — find the words for what you feel

You know what you feel. Finding the words is another matter. SomaMap is a vocabulary and journaling tool that helps you describe a physical sensation in your own words and keep those words somewhere you can find them again.

You tap where you feel it, describe it in a sentence or two, and answer three short questions about how it feels — spot or spread, steady or wave-like, what it reminds you of. A Muse Spark Open model reads what you wrote and chooses which of SomaMap's own reviewed questions comes next, and offers a handful of plain words you may tap or ignore. The wording of every question comes from a fixed, reviewed bank: the model selects, it never writes the questions, and the region on the card comes from the app's own fixed body map.

What comes out is a card with exactly five fields — where, the words you chose, what it feels like, how long, and your own sentence — and nothing else. There is no field for what a sensation might mean, because the schema has no such field: a model reply carrying a sixth key is rejected whole and reviewed fallback wording stands in. The journal keeps your cards and counts the words on them, and says plainly what those numbers are: counts of your own words, and nothing else.

SomaMap does not diagnose, interpret or advise, and it never says what a sensation means. Every screen says so. A check-in works without an account; signing in is what keeps a card.

Live at https://somamap-ebon.vercel.app
https://github.com/tobacreator/somamap

## More from Meta Global AI Developer Hackathon

- [Wildseek](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4292)
- [KnowMe](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4293)
- [LawnLens](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4294)
- [Cruyff](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4296)
- [MolAstria](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4297)
- [SAMREEN.J](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4298)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/4295. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt