QuantDinger is a self-hosted AI trading platform for independent traders, Python strategy authors, and small teams. It carries a trading idea through the full loop: AI market research, Python strategy code, server-side backtesting, paper or live execution across crypto exchanges and traditional brokers, and monitoring. Market data, strategy code, broker credentials, and deployment stay under the operator's control; it is not a signal service.
The v5 backend separates the HTTP API from long-running work: trading, scheduling, Celery jobs, and migrations run as distinct processes, with Celery handling finite retryable jobs and a trading worker owning long-lived strategy runtimes. State lives in PostgreSQL, with separate Redis instances for cache and durable jobs, and an optional overlay adds Prometheus, Grafana, and Alertmanager. Strategy ownership uses leases, heartbeats, and fencing tokens, and broker credentials are encrypted at rest.
Agents reach the platform through an Agent Gateway under /api/agent/v1 and a bundled MCP server, so clients such as Cursor, Claude Code, and Codex can call approved tools without holding broker credentials or admin JWTs. Agent tokens are hashed, scoped, rate-limited, and audit-logged; agent trading is paper-only by default, and live orders require a trading-scoped token with paperonly=false, a server-side AGENTLIVETRADINGENABLED flag, and operator-configured limits and allowlists.
| Area | Surface |
|---|---|
| Indicators | Python chart overlays, markers, bands, and signals |
| Strategies | Strategy API V2 intents, sizing, risk, backtests, and live runtime |
| Crypto | Binance, OKX, Bitget, Bybit, Gate, HTX, and adapter extensions |
| Traditional brokers | IBKR and Alpaca workflows |
| AI providers | OpenRouter, OpenAI-compatible APIs, Google, DeepSeek, Grok, MiniMax, and custom endpoints |
| Automation | Human API, Agent Gateway, MCP server, Celery jobs, schedules, and notifications |
Features
- AI research: multi-provider market research and analysis feeding strategy generation
- Strategy API V2: Python strategies expressing intents, position sizing, and risk rules that run in both backtests and live runtime
- Backtesting: server-side backtests and experiment workflows
- Paper and live execution: crypto exchanges and brokers behind one runtime, with live trading gated behind explicit enablement
- Agent Gateway and MCP: scoped, audited agent tokens; MCP tools for Cursor, Claude Code, and Codex
- Clients: desktop web, mobile H5, and a human API, each proxied through Nginx
- Process separation: API, trading worker, scheduler, Celery, and migrations as independent containers
- Observability overlay: JSON logs, request IDs, Prometheus metrics, Grafana dashboards, and alert rules
- Hardened production overlay: non-root containers, read-only root filesystem, dropped capabilities, resource limits, loopback-only ports
- Credential security: encrypted broker credentials and MFA secrets, production config validation script
