Where finance, data and AI converge — and every order is earned, not sent.

No. The AI is an advisory and reasoning layer. It cannot bypass the Risk Engine, and it cannot turn generated text into an executable order. Position sizing, exposure limits, drawdown checks, margin checks and order validation are all owned by deterministic services. Any model output that would become an action is schema-validated and allowlist-checked first.

TradingView indicators and strategies, MetaTrader 5 indicators and expert advisors, cTrader indicators and cBots, AI tools and automation products. Each listing declares a compatibility matrix, an immutable version ID, a code hash, a changelog and a verification status before it is published.

Connector credentials are encrypted at rest and are never exposed to the browser. We prefer least-privilege, trade-only credentials, IP restrictions where the provider supports them, and separate keys per environment and per account. Secrets live in a secrets manager, never in source code or frontend configuration. The desktop agent stores local secrets in the OS keychain and never uploads raw broker secrets to the cloud.

No, and we say so explicitly. Backtests record the exact data source and version, timeframe, fees, slippage, spread and latency assumptions, strategy version and parameter set so results are reproducible — but reproducibility is not predictability. Historical and backtested performance is never a guarantee of future results. Trading involves risk of loss and this product does not provide investment advice.

No. The default state is off. Activation requires an explicit action from you plus a visible risk profile. You get per-account enable/disable, per-strategy enable/disable and a global emergency stop. Emergency stop prevents new orders immediately; closing existing positions is a separately controlled action because forced liquidation carries its own risk.