beanquery-mcp

Overview

The Beanquery MCP Server bridges AI assistants with Beancount financial ledgers through the Model Context Protocol. It exposes tools for setting ledger files and running BQL queries, plus resources for accessing account and table information. This experimental implementation enables AI to analyze financial data in Beancount format, though it's still evolving and may undergo significant changes.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• set_ledger_file

  Set the Beancount ledger file to use for queries.

  Try:Set the ledger file to /path/to/my/ledger.beanTry:Use sample.bean for queries
• run_query

  Run a BQL query against the loaded Beancount file.

  Try:Show all transactions from January 2023Try:Calculate total expenses this year

Comparable tools

Installation

mcp dev server.py

uv run mcp install server.py --with beancount --with beanquery

uv run mcp install server.py -v BEANCOUNT_LEDGER=/path/to/your/ledger.bean --with beancount --with beanquery

{
  "mcpServers": {
    "beancount": {
      "command": "uv",
      "args": ["run", "mcp", "install", "server.py"],
      "env": {
        "BEANCOUNT_LEDGER": "/path/to/your/ledger.bean"
      }
    }
  }
}

FAQ

Is this server production ready?

No, this is an experimental implementation that may undergo significant changes. It's recommended to use it in a development environment.

What data privacy concerns should I be aware of?

This tool may transmit parts of your Beancount ledger to third-party LLM services. Use caution with sensitive financial data, consider masking information in test ledgers, and prefer self-hosted LLMs when possible.

Compare beanquery-mcp with