Zweispace vertical AI token family

MAWASHI

Sumo AI matching for wrestlers, coaches, stables, events, JavaScript demo fights, and robot physical intelligence.

Team context Wrestler-led knowledge layer, public-safe until approvals
First demo Browser JavaScript, no Unity dependency for v0.1
Robot PI Balance, grip, collision timing, ring-out recovery

Standalone target

zweichain.com/prj/mawashi/ can become the public MAWASHI home.

MAWASHI follows the Zweispace vertical-AI pattern around expert systems such as AI税理士 and AI中小企業診断士, then moves that logic into embodied sumo intelligence. BOX stays the boxing reference project. MAWASHI becomes the sumo branch: matching, approved wrestler knowledge, dohyo simulation, and robot physical intelligence.

MAWASHI system

One vertical, four linked surfaces.

Public wording stays wrestler-safe until names, records, stables, photos, and endorsements are approved.

01

Rikishi graph

Capture approved technique notes, tachiai patterns, grip preference, ring sense, and coaching knowledge.

02

Training matching

Match wrestlers, coaches, stables, clubs, practice partners, clinics, and events by safety-first constraints.

03

JS dohyo demo

Use browser JavaScript for tachiai, grip contest, ring position, style scoring, and explainable demo receipts.

04

Robot PI

Use sumo motion as a high-signal physical-intelligence domain for balance, push force, and recovery.

AI matching pilot

Match by style, safety, and training goal.

The v0.1 matcher is a transparent browser mock. Real deployment needs consented profiles and human review.

Recommended route

Wrestler-led clinic pilot

Set a goal and score the first safe matching route.

0 / 100

Playable MVP

Zantei yokozuna vs Challenger A/B/C.

The fight runs locally in browser JavaScript. It is a demo model, not real athlete evaluation.

Challenger setup
Zantei Yokozuna holds the title Pick Challenger A/B/C and run the JavaScript fight.
Champion 78
Challenger 0
Technique ready
The dohyo is ready.

Token track

Product first. Utility token only after review.

MAWASHI can later use token receipts for matching credits, demo-fight records, model challenges, and community participation. The public story should stay clear: no investment promise, no gambling framing, and no athlete access claims.

Token guardrails

Build path

From BOX reference to sumo-specific matching.

1

Approve public story

Confirm wrestler/team wording and PR Times reference before public release.

2

Profile schema

Define consented wrestler, coach, stable, club, event, and sponsor fields.

3

Model arena

Adapt the BOX browser arena into a MAWASHI style and safety scoring challenge.

4

Robot bridge

Translate balance, foot pressure, and push recovery into physical-intelligence datasets.