mcp -tradersis

This MCP server provides tools to fetch, analyze, and visualize stock prices and financial statements for quantitative trading research. It exposes capabilities to retrieve real-time and historical price data for single or multiple tickers, extract income statements, balance sheets, and cash flow data, and compute key financial and technical metrics such as PE ratios, ROE, and moving averages. The outputs are designed to be JSON-serializable for easy consumption by LLMs and visualization tools. To use it, configure your MCP client (for example Claude or other LLM environments) to connect to the MCP server endpoint and call the available tools by name. The README example demonstrates wiring the server into Claude’s tool directory so that “tradersis-mcp” appears as a selectable tool. Once connected, you can ask questions like “What is the ROE of AAPL?”, “Show me the last 6 months of MSFT price data,” or “Plot buy/sell signals for TSLA,” and receive structured data and visualizable outputs. The server is designed to return information that can drive charts, tables, and concise trading insights.

Prerequisites:

• Python 3.8+ installed on your system
• Access to a shell/terminal
• Git installed

Installation steps:

1. Clone the repository: git clone https://github.com/Nan-Shen/mcp-server-tradersis.git cd mcp-server-tradersis

2. Set up a Python virtual environment: python3 -m venv .venv source .venv/bin/activate

3. Install dependencies and install the MCP runtime (following the README’s approach). If the project uses uv (the MCP runtime) for Python, install or install via your preferred method (examples shown). Then install project dependencies:

   • If using uv: uv sync
   • If using pip requirements: pip install -r requirements.txt

4. Run the MCP server via the configured command (example based on the README): uv run --directory /directory/to/your/mcp main.py

5. When running, ensure your MCP client is configured to point to the server. For Claude, add a new MCP server with the JSON configuration shown in the README to expose the tradersis-mcp tool.

Notes:

• Replace /directory/to/your/mcp with the actual path to the MCP server code (where main.py resides).
• If you are using a container or a different runtime, adapt the command accordingly (see mcp_config for the available command and args).

Tips and caveats:

• Ensure network access between the MCP client (LLM) and the Python MCP server so that tool calls can be executed and responses returned in JSON.
• The server outputs are JSON-serializable for easy ingestion by LLMs and charting tools. When building prompts for LLMs, prefer structured outputs (objects and arrays) to enable visualizations.
• If you update main.py or the directory structure, restart the MCP server so the client can discover the new capabilities.
• If you encounter path issues, verify that the --directory path points to the folder containing main.py and related modules.
• Environmental variables can be used to configure API keys or data sources (e.g., stock data providers). Add them to the env object in mcp_config if your deployment requires them.