A private terminal for perp markets. It watches where the crowd is positioned and what gets force-unwound, reads the risk in plain terms, and lets an agent act on it, on your terms.
Raw positioning is noise until something acts on it. intenthybrid runs the same three steps every time, so a crowded book becomes a decision instead of a chart.
Crowding index and liquidation flow across perp markets, moving as positioning shifts. One feed for where risk is building.
The AI turns raw positioning into a plain risk read: where the crowd sits, what unwinds if it breaks, how crowded is too crowded.
Set an alert, or let a scoped agent execute inside hard limits you define. Risk you can act on, not just look at.
Ask about any market in plain language. The conversation lives in your browser, prompts run on infrastructure you control, and nothing is stored on our servers.
Every read is an endpoint an agent can pay for over plain HTTP. No keys, no accounts, no monthly invoice. Just a few cents of USDC, settled in under a second.
Build a scoped agent that watches the signal and executes only inside the hard limits you set. It acts when conditions hit, and never beyond what you allow.
Track crowd positioning and forced unwinds across perp markets at a glance. Sort by where the squeeze is forming, and open any market for the full read.
How a read moves through intenthybrid. Market data and your own rules feed a left to right spine that detects crowding, reads the risk, and acts, with privacy and pay-per-call wrapped around the whole thing.
intenthybrid is a private terminal for perp-market risk. It reads where the crowd is and what unwinds, then puts the decision one keystroke away, by chat, by alert, or by agent.
A small stack with sharp edges: private inference, a crowding engine, pay-per-call settlement, and scoped onchain execution.
Built in the open and shipped in phases. No dates promised, just the order of work and what is live today.
What you can use today.
Being built right now.
Lined up after the above.
Further out, not committed.