The parts every trading agent rebuilds, already built
Schedule work, clear a batch, score a fill, verify an enclave, check a signature. All native, all for fractions of a cent.
Agent functions are small, well-tested operations the network provides to every contract, program and agent. Use them as you would a standard library, and spend your time on the strategy instead.
Every team that builds a trading or procurement agent ends up writing the same scaffolding. A cron job that fires on time. A routine to clear a batch at a fair price. A score for how good a fill was. A check that a result really came from a secure enclave. On Paxeer X those are built into the network. You call them from a contract, a program or an agent, and each costs a fraction of a cent.
What is in the box
An execution desk, assembled from parts
- 01
Collect
An agent gathers client orders for the hour, each priced as an offset from the reference.
- 02
Clear
At the top of the hour the scheduler fires. The clearing function returns one price for everyone, so no client is picked off.
- 03
Score and pay
Each fill is scored against the reference. Execution agents are paid by streaming payment, with the rate tied to their score.
- 04
Prove
The whole run is a chain of receipts, so a client who asks why they got that price gets an answer that checks out.
Why native beats a library
A library runs on your server and dies with it. A native function runs inside the transaction, at the same moment the money moves, with the same receipt. The scheduler keeps firing while you sleep, the clearing price is computed from committed data, and the attestation check cannot be skipped by a hurried deploy.
Mix them with streams, budgets and escrow. A scheduled job can settle a stream, a scored batch can release an escrow, and an attested worker can be paid per call.
Common questions
Related
Build your first strategy
Schedule a job, clear a batch and let the network do the chores.

