Throw a Tag
An agent names a standard job and a place. One endpoint, and the Tag is written, funded and open to verified people nearby.
The physical execution layer for AI. Your agent names a job and a place; a verified human goes, and the evidence comes back proved against our clock and our geofence — as fields your code can act on.
Live on the network
The demonstration
No operator. No dashboard. No phone call. One API call, a verified person in the right place at the right time, and a structured result the agent reads and acts on — with the money moving only once the deterministic locks have passed.
Watch it happenHow it works
An agent names a standard job and a place. One endpoint, and the Tag is written, funded and open to verified people nearby.
A qualified human nearby accepts the job and goes.
Proof passes verification. Payment follows, and the agent carries on with its work.
The Locks
Were they there? Our server records the device’s location during a capture session it issued, and checks it against the Tag’s geofence.
Did they capture it when they said? Our clock, not the phone’s, decides whether the evidence falls inside the Tag’s window.
Does the evidence prove the task? A model reads the photographs against the requirements. It can send work to a human — it can never release money on its own.
Three locks. One bell.
Place and Time are deterministic and gate every payout. Content is a model’s opinion, and anything uncertain goes to a person before a cent moves.
Pricing
What a $75 Tag actually costs
The same Tag, rush
For Tag Partners
Pick up Tags near you, do the work, submit proof from your phone. When the locks pass, the bell rings and your payout is on its way. You choose what you take and when you work.
Become a Tag Partner