Your code is
your capital.
Build a trading agent, prove it on a standardized simulated market, and get funded on results — with a track record nobody can fake and zero capital of your own.
We don't promise profit. We measure. Free practice arena, no card, no capital.
$0
Capital required
70%
Profit split
1,828
Journal entries
4
Agents on record
Built so nobody has to be trusted.
Including us. Honesty here isn't a value statement — it's the architecture.
Records nobody can fake
Every order, fill, and verdict is sha256-chained in an append-only journal. A retroactive edit — ours or yours — breaks the chain in public. Anyone can verify any agent over the open API.
GET /v1/verify/:agent
A simulator that plays fair by being harsh
You pay the spread, size-dependent slippage, and taker fees. Nothing fills on stale data, and orders never see the future. Fills are never better than the real feed.
parameters public & versioned
The odds, published
Prop firms hide their pass rates. Ours are on a public page, updated live, failures included. When rules change, the version changes — old records keep theirs.
see /stats
An agent is one function.
Scaffold a project, implement decide(), run. Indicators, ML models, LLMs — anything goes, and it all runs on your machine. We never see your code; the market grades it.
Scaffold
One command creates a working agent. Your strategy is a single Python function.
Trade
It runs on your machine against a standardized simulated market. Practice is free.
Prove
Your track record accrues in a verifiable public journal.
Get funded
Pass a challenge and keep 70% of the profit on a funded simulated account, paid in USDC.
# your entire strategy — the SDK does the rest
class MyAgent(StrategyAgent):
def decide(self, market):
if market.position == 0 and market.change_pct(60) < -0.3:
return "buy"
if market.unrealized_pct and market.unrealized_pct > 1:
return "close"The market doesn't care about your pitch — it cares about your math.
Connect an agent in ten minutes. Beat the do-nothing baseline. Then come back for the challenge.