A multi-strategy execution engine that sizes every position from its stop, refuses any trade that breaks a limit, and writes down why. Seven strategy families, a four-level circuit breaker and a double-entry ledger — running on hosting you already own.
Free tier runs the full engine in paper mode. No card required.
This installation has not traded yet, so there are no performance figures to show. Numbers appear here as the engine builds a record — nothing on this page is illustrative or placeholder data.
The order is deliberate. Open risk is managed before new risk is taken — always, without exception.
Candles and quotes are pulled for every active instrument and timeframe, then written once so every strategy reads the same history.
Open positions are marked to market. Stops, targets, breakeven moves and trailing stops are applied before anything new is considered.
Each pool is measured against its daily loss limit and drawdown ceiling. Breakers escalate through four levels and can flatten the book.
Only then do strategies look for entries — and every candidate must clear fourteen risk checks before it reaches a broker.
Most losing accounts pick a lot size first and place the stop wherever it fits. This engine inverts that. You state the risk you accept; the strategy states where it is wrong; the size falls out of the arithmetic.
Widen the stop and the position shrinks. The money at risk does not move. That single property is what turns a run of losses into a survivable drawdown instead of a blown account.
Breakers escalate on their own and, at the top level, need a human to clear them. A system that can talk itself back into trading after a 15% loss is not a risk system.
Position size halved for the rest of the session.
No new positions. Existing ones are managed to the exit.
Everything is closed. Trading halts for 24 hours.
Account locked. Only a person can restart it.
Daily loss, drawdown, exposure, correlation, spread, margin, session, news proximity, duplicate signals and more. A refusal is recorded with its reason so you can see what the system stopped and why.
Three long positions in EUR/USD, GBP/USD and AUD/USD are close to one position at triple size. Correlated exposure is measured and capped across the whole book, not per pool.
Every deposit, fee, commission and realised profit posts balanced journals. The trial balance either proves out or the system tells you it does not.
Automation is not an edge. This platform executes rules faster and more consistently than a person can, and it stops when the rules say stop. It cannot make an unprofitable strategy profitable.
A good backtest proves very little. Fitting parameters to past data until the curve looks good is the most common way traders lose money. The walk-forward tool here exists to show you how much of an edge survives on data the fit never saw — and it will often tell you the answer is "almost none".
The default feed is simulated. Out of the box the engine runs against a synthetic market so you can exercise every feature safely. Live trading needs real broker credentials, and results there will differ.
Taking other people's money is regulated. The pooling and payment features are built to a professional standard, but operating them for third parties requires licensing in most jurisdictions. That is your responsibility, not the software's.
Paper trading with the full engine.
Live venues and the complete risk framework.
Multi-account operation and full auditability.
Deploy to your own hosting in a few minutes. Run it on paper for as long as you like.