emerald-exchange¶
Unified Finance MCP Server + A2A Agent for the agent-utilities ecosystem — fully abstracted exchange backends for equities, crypto, derivatives, and prediction markets, with pre-trade financial hardening.
Official documentation
This site is the canonical reference for emerald-exchange, maintained
alongside every release.
Overview¶
emerald-exchange is a single multi-asset finance hub that exposes execution,
market data, risk, signals, fundamentals, and wallet intelligence as typed,
action-routed MCP tools. Every order is validated by a pre-trade risk guard
(OS-5.1 financial hardening) before it reaches an exchange backend. It provides:
- Five exchange backends — Paper (the always-on default), Alpaca, CCXT
(Binance / Coinbase / Kraken and 100+ venues), Freqtrade, and Polymarket — all
behind one
ExchangeBackendprotocol. - Action-routed MCP tool domains — market data, orders, portfolio, risk, signals, strategy, crypto, derivatives, market-making, statistical arbitrage, prediction markets, fundamentals (SEC EDGAR), and wallet intelligence.
- Financial hardening — paper-first default, Kelly-criterion position sizing, circuit breakers, regime-shift detection, and an instant kill switch.
Two operating invariants: trading defaults to paper and never to live, and all quant compute is delegated to the Rust epistemic-graph engine rather than re-implemented in Python.
Explore the documentation¶
- Installation — pip, source, extras, and the prebuilt Docker image.
- Deployment — run the MCP and agent servers, Docker Compose, Caddy + Technitium.
- Usage — the MCP tools, the Python API, and the cockpit CLI.
- Overview — enterprise features, tool surface, and architecture.
- Configuration schema — the
tradingconfig block and backend matrix. - Concepts — the
CONCEPT:EE-*registry.
Quick start¶
Run it as a network server with a published port:
See Installation and Deployment for the full matrix (PyPI extras, Docker image, every transport, the A2A agent server, reverse proxy, and DNS).