This MCP server provides a free, open-source financial analysis and quantitative trading environment. It exposes a departmental MCP architecture that integrates traditional indicators, price action analysis, an economic calendar, fundamentals, and news data, and is optimized for interaction with large language models and automated trading workflows. You can run the server locally with a single Python command and then connect LLMs or client processes via the MCP interface (RPC/PubSub) to request analyses, run trading strategies, or query market data. The included tooling is designed to simulate a real financial firm’s operations, enabling data flow between departments and AI agents in a cohesive, testable environment.

To start, run the MCP server with: python -m finance_trading_ai_agents_mcp (or use your environment’s .env file to configure credentials). The server will expose structured financial data and functions that your MCP clients can call, such as technical indicators, calendar events, fundamentals, and news feeds. You can also leverage the provided examples and env_example.toml to customize the setup. The project is designed to work smoothly with AI assistants and trading bots, enabling real-time data streaming and cross-process communication via local RPC/PubSub channels.

Prerequisites

• Python 3.8+ (recommended latest available)
• pip (comes with Python)
• Access to the internet to install package dependencies

Installation steps

1. Optional: clone the repository or install from PyPI

• From PyPI (recommended): pip install finance-trading-ai-agents-mcp

• From source: git clone https://github.com/aitrados/finance-trading-ai-agents-mcp.git cd finance-trading-ai-agents-mcp pip install -r requirements.txt

  If you want an editable install during development:

  pip install -e .

2. Prepare environment variables (example):

• Create an .env file or export them in your shell: export AITRADOS_SECRET_KEY=your-secret-key-here

  Add any other required configuration keys as described in docs

3. Run the MCP server

• If installed via PyPI: python -m finance_trading_ai_agents_mcp
• If running from source after installation: python -m finance_trading_ai_agents_mcp

4. Verify startup and access docs

• Check console output for startup confirmation and endpoints
• Refer to the Quick Start docs for how to connect MCP clients

Notes and tips:

• The server relies on AITRADOS secret keys for data access; ensure AITRADOS_SECRET_KEY is set in the environment.
• The MCP supports local RPC/PubSub communication, enabling cross-process interaction and multi-language data sharing.
• If you plan to customize MCP behavior, check the example env and config.toml in the project repository for guidance.
• For production deployments, consider running behind a local reverse proxy and configuring secure credentials and access controls.
• Common issues include missing dependencies or environment variables; use a dedicated .env file and confirm all required vars are present before starting the server.