Account
Connecting to OANDA...
Portfolio Hub · Account Overview
Account performance
Read-only execution scope
Actual OANDA executions
Choose one environment and one validated connection. Practice and Live are never combined.
Loading Practice executions…
Actual-execution realized P/L
-- -- metric executionsSelected-account context
OANDA account snapshot
Balance and NAV are account context. They are not bot execution performance.
- Balance
- --
- NAV
Bot executions
Exact bot subsets from the selected execution ledger.Execution ledger
Trade Center Activity
Cumulative realized P&L
Monthly realized P&L
Portfolio Level · Execution Operations
Trade Center
Poller health, account exposure, strategy decisions, and broker-confirmed trades.
Select a Trade Center bot
Strategy execution
Choose a bot to inspect its latest completed OANDA candles.
Strategy chart
Completed OANDA candles with the Python strategy’s channel, EMA, signals, and frozen brackets.
Latest decision
--Risk gates
--Basket order management
Read-only normalized worker evidence. No manual stop or close actions are available here.
TP1 milestone confirmed keeps 100% open and advances the whole protective stop to +0.5R. TP3 (3.5R), then TP2 (2R), then the 30-bar stale exit are mutually exclusive terminal close intents evaluated on completed H4 candles. A broker protective-stop fill is authoritative intrabar and reconciles before completed-candle lifecycle work.
Instrument states
Independent order, risk-reservation, lifecycle, and recovery evidence for each basket child.
Completed-H4 cadence
Oldest unresolved boundary first; the cursor advances only after every instrument outcome is terminal.
Immutable parity & runtime attestation
Normalized evidence timeline
Cadence, decision, admission, order, lifecycle, reconciliation, and risk events from the versioned DTO.
Decision ledger
Every completed-candle outcome, including no-signal and blocked evaluations.
Trade history since activation
Broker-confirmed Trade Center trades only, newest closure first.
Saved vs applied execution policy
Read-only comparison of the immutable saved version and the active native-worker row.
--
--
Portfolio Hub
Your portfolios
Organize exact saved strategy versions. An account reference does not reserve or deploy.
Open Portfolio Hub to load your portfolios.
Draft includes unbound and inactive account references. Active includes deployment-linked records, including paused or needing review.
Existing strategy deployments
Manage portfolio
A copy keeps eligible saved strategy references. Account binding, reservations, orders, execution history, P&L and research evidence stay with the original.
This records a reference only; it does not reserve the account or prove it is available for deployment. An account occupied by another deployment cannot be used for a new deployment.
Saved strategy components
Choose one exact saved version for each selected bot.
Exploratory research
Portfolio Lab
Simulate selected saved versions on one shared account NAV, margin, and simultaneous-position ledger.
Choose saved strategies and versions for a new research run. This does not change a portfolio's saved or deployed composition.
Choose an owned Practice connection for research. Running research does not deploy or trade.
No research result selected. Current portfolio and research source choices are shown above.
Blended Equity
Blended Drawdown
Selected Strategy Contribution
Admission & Opportunity Summary
Rejected and reduced opportunities are isolated counterfactuals. They are not additive portfolio returns because admitting one trade changes capital available to later trades.
Portfolio Trade & Rejection History
Audit history
Strategy Version Archive
Reversible audit-only history. Archiving changes selector visibility only; immutable source, results, evidence, and hashes remain untouched.
Strategy Hub
Your strategies
Open a bot to review its executions, research saved versions, or manage deployment.
Loading your strategies…
Strategy Hub · Bot workspace
Select a bot
Open a strategy from the Hub.
Actual-execution realized P/L
-- -- metric executions · all executed versionsExecution ledger
Exact bot ID subset from the selected ledger.Bot Activity
Cumulative realized P&L
Monthly realized P&L
Bot Stats
Price Replay
Candles, entries, exits, trade paths, and scoped deviation markers.Price overlay controls
Backtest Equity Curve
Drawdown
Rolling Metrics
Live vs Backtest
Deviation Check
Backtest Trade History
OANDA Candle Regime
Market Regime Performance
ATR(14) volatility is bucketed into trailing 300-bar terciles; ADX(14) above 25 is trending, otherwise ranging. Trades are tagged at entry.
Single-Run Analysis
Trade Direction Breakdown
Compares win rate and average P/L for the long and short trades this backtest already produced. No additional run is needed.
Session Breakdown
Quant Tools
Sweep, walk-forward, and Monte Carlo checks.Parameter Sweep
Parameter Stability
Fixed PF scale: red at 0.6 and below, green at 1.4 and above.
Performance Clusters
Walk-Forward
Attributed strategy registry
Community
Discover creator-published bot metadata and validation status. Strategy source stays visible only to its creator; adoption is unavailable in this beta.
Saved Configurations
Select at least two immutable configurations.
Workspace Recycle Bin
Deleted strategies and portfolios
Bot Builder
Step 1
Start from your own logic
Write one BaseStrategy class that produces a signal, stop, and target for one instrument. The editor begins with a small structure you can replace with your own entry and bracket rules.
Step 2
Choose the execution model
Instrument availability and order limits come from your selected OANDA account.
Step 3
Name and shape the bot
Data Models
Write your own signal and bracket rules in the editable Python source.
Step 4
Create the draft bot
Create a Trade Center-compatible bot here, then backtest from its own workspace. Nothing becomes live from this page.
Draft Created
Draft bot created
Open its workspace to backtest, edit code, and promote deliberately when ready.
