Composer MCP Server

The Composer MCP Server bridges the gap between conversational AI and real‑world investing by exposing a rich, domain‑specific API that lets language models validate, refine, and execute trading ideas. Rather than hard‑coding strategies or manually backtesting data, an AI assistant can now ask for a strategy that meets specific risk constraints, request a backtest against a benchmark, or even trigger live trading of multiple portfolios—called symphonies—in parallel. This end‑to‑end workflow empowers developers to prototype sophisticated investment logic directly through natural language prompts, accelerating the iteration cycle from hypothesis to deployment.

At its core, the server offers a set of intuitive resources: strategy creation, backtesting, search and recommendation, performance monitoring, and, for subscribers, live portfolio management. Developers can compose strategies using standard technical indicators such as RSI, MA, and EMA across a broad universe of equities and cryptocurrencies. Once defined, the same model can instantly run a high‑performance backtest, compare results to market indices like the S&P 500, and visualize performance—all without leaving the chat. The search endpoint further lets an AI scour a library of 1,000+ pre‑built symphonies to find one that matches user criteria, making it trivial to discover niche strategies or benchmark against peers.

Real‑world use cases abound. A fintech startup can let its customer support bot answer “Which crypto strategy has a maximum drawdown under 30%?” and receive a ready‑made portfolio that the user can deploy. An algorithmic trader might ask the assistant to “Rebalance my equity symphonies weekly based on recent earnings data,” and have the server execute those adjustments automatically. Portfolio managers can query “Identify my best‑performing symphonies and explain why” to surface insights that would otherwise require manual data crunching.

Integration is seamless with any MCP‑enabled client—Claude Desktop, Claude Code, Cursor, n8n, or custom LLMs. By simply adding the Composer connector, developers unlock a powerful toolkit that turns conversational prompts into actionable trading commands. The server’s design prioritizes speed (fast backtests), transparency (exposed metrics and plots), and extensibility (custom indicator support), giving it a distinctive edge over generic financial APIs. In short, Composer MCP turns AI into an intelligent, data‑driven trading partner that can think, test, and trade in real time.