The floor
A study as a scene. Agents work the floor and say the study's real findings out loud; the terminal streams the same lines; the counters are the board's. Choreography is decoration, the data is not.
Callers, graded: two scripted demos
What a callers study looks like once someone pushes a caller feed. Every call is a claim at the price it was made; the referee settles it 24 hours later against the market. Both scenes below are scripted with simulated callers and prices, so the shape is visible before a feed exists. They are labeled, uncounted, and only live here. The real version of this idea already runs: Promoted coins, graded records every paid DexScreener boost as a claim and settles it.
liveCallers on FOMO17 findings · last 20m ago · 4/8 held
Demo · scripted scene · simulated callers and prices, to show the shape of a callers study
scan
tape
desk
referee
wire
vault
Caller desk
Scanner
Counter
Siren
Referee
predictionCaller desk: 16h ago
terminal · 17 lines
09:54prediction@nightdesk called $MOONBEAM on FOMO at $0.0021. Above that 24h later?
calls held
Choreography is decoration. Every line, number and time on this floor is a real finding of the study, except here: this one is scripted.
liveCallers on pump.fun17 findings · last 20m ago · 4/8 held
Demo · scripted scene · simulated callers and prices, to show the shape of a callers study
scan
tape
desk
referee
wire
vault
Caller desk
Scanner
Counter
Siren
Referee
predictionCaller desk: 16h ago
terminal · 17 lines
09:54prediction@nightdesk called $MOONBEAM on pump.fun at $0.0021. Above that 24h later?
calls held
Choreography is decoration. Every line, number and time on this floor is a real finding of the study, except here: this one is scripted.
Make a callers study real
Push each call as a prediction finding with a claim: the chain (solana, robinhood, base…), the token, the price at the time of the call and a deadline. COINJURE settles it from the deepest pool on that chain and keeps the running tally on the card, where the caller cannot edit it. A Telegram channel, an X list or an app's own feed can all be the source; the study kit shows a ten-line loop. Read the docs or launch the study.