This Ghostfolio MCP Server provides a programmable interface to Ghostfolio data, allowing read and write access to portfolios, holdings, transactions, and market data through a modern MCP API. With the available tools, you can retrieve portfolio performance, holdings, dividends, and allocations, as well as manage accounts, import transactions, and monitor orders and balances across accounts. The server also supports security features such as read-only mode, rate limiting, and optional Sentry integration for observability. To get started, configure the required Ghostfolio URL and API token via environment variables, then run the Python-based MCP server. Once running, you can invoke operations like get_portfolio_performance, get_portfolio_holdings, get_accounts, and create_account through the MCP API, and you can enable read-only mode to prevent any write actions during monitoring.

Prerequisites:

• Python 3.11 or higher
• Access to a Ghostfolio instance
• Ghostfolio API token

Installation steps (recommended):

1. Install the MCP server from this repository or via PyPI:

   Using UV (recommended)

   uvx ghostfolio-mcp

   Or using pip

   pip install ghostfolio-mcp

2. Set up environment variables for your Ghostfolio connection. Create or edit a .env file or export variables directly:

   export GHOSTFOLIO_URL=https://domain.tld:3333 export GHOSTFOLIO_TOKEN=your-ghostfolio-token export GHOSTFOLIO_VERIFY_SSL=true export GHOSTFOLIO_TIMEOUT=30

   Optional: enable read-only mode

   export READ_ONLY_MODE=false

   Optional: rate limiting and logging

   export RATE_LIMIT_ENABLED=false export LOG_LEVEL=INFO

3. Run the server:

   Using Python (as configured in this repository)

   python run_server.py

   Or via installed script (if you installed via PyPI)

   ghostfolio-mcp

4. (Optional) Development setup:

   Install from source and install dependencies

   git clone https://github.com/mhajder/ghostfolio-mcp.git cd ghostfolio-mcp uv sync --group dev

5. Verify the server is running by hitting the MCP endpoints or checking logs. Ensure the Ghostfolio URL and token are valid to avoid authentication errors.

Tips and known considerations:

• If you enable READ_ONLY_MODE, all PUT/POST/DELETE operations are disabled, making it safe for monitoring dashboards.
• Configure MCP_TRANSPORT (stdio, sse, or http) based on your deployment scenario. For local usage, stdio is typical; for streaming HTTP clients, consider sse or http transports.
• Secure your MCP HTTP endpoint if exposed externally by setting MCP_HTTP_BEARER_TOKEN and enabling SSL where applicable.
• Sentry integration is optional. If SENTRY_DSN is not configured, the server runs without error tracking.
• Use the provided environment variables to fine-tune timeouts, SSL verification, and logging to suit production workloads.
• For best results, ensure your Ghostfolio instance is accessible from the MCP server host and that the API token has the necessary permissions for the requested operations.