An AI agent workspace for real trading strategies. It helps traders and strategy builders develop Pine Script strategies, test how they performed, compare expected signals with real alerts, and review broker fills when broker data is connected.

Generators stop at code. The Algo Agent keeps testing: it runs baseline backtests, finds weak spots, proposes changes, compares strategy versions, and saves a research record so you can see why a strategy was promoted or rejected.

No. The platform helps you research, validate, and audit trading strategies. You stay in control of any broker or automation you use, and Alerts vs Fills shows what actually happened after signals were sent.

Yes. Full platform access with no credit card required during beta. Beta users keep lifetime early-adopter pricing when paid plans launch.

Paid plans include monthly AI credits (Basic 200, Professional 500). Beta credits are issued manually by admins. Credits pay for Algo Trade Analytics-hosted AI work: AI Pine editing, Algo Lab research, and MCP-started agent runs (no BYOK). External AI clients may have separate costs. Monthly credits reset each cycle; top-ups roll over.

During beta, approved users may receive MCP access while packaging is finalized. After beta, full MCP access is planned for Professional. Basic keeps the in-app Algo Lab workflows; Professional adds external AI agent orchestration so Codex, Claude Code, Cursor, ChatGPT, or another MCP-capable client can run research, gates, and backtests through Algo Lab.

Start in Algo Lab with a Pine strategy or a new idea. Let the agent test versions against historical bars, then connect alerts or upload CSV data to check whether real signals match the backtest. Link a broker only when you are ready to review fills.

No. You can start with Algo Lab and Backtest vs Alerts without connecting a broker, as long as market data is available in your workspace. A broker connection is only needed for fill review, broker fill markers, and portfolio data.