MCP Finance Assistant
Connects Notion databases with LLMs to manage and analyze personal finances through natural language queries and bank statement uploads.
README
MCP Finance Assistant
Description
MCP Finance Assistant is a modular, terminal-based financial assistant that connects with Notion and can leverage LLMs (OpenAI, Mistral, Anthropic) to help you manage, analyze, and track your finances with ease.
It uses a client-server architecture: the client is built in Python (Textual UI, LangGraph agents), and the server (Node.js/TypeScript) acts as an MCP tool provider, integrating with your Notion databases.
Features
- Terminal UI (TUI) for a fast, distraction-free experience.
- Modular, graph-based agent design (LangGraph, LangChain).
- Native Notion integration for storing and querying accounts & transactions.
- Extensible with OpenAI, Mistral, or Anthropic for advanced AI responses.
- Clean separation of client (Python) and server (Node/TypeScript, MCP).
- Easy configuration with environment variables.
📸 Demo GIFs
See MCP Finance Assistant in action! Below are sample use cases you can try, with animated GIFs for each scenario.
1. Uploading a Bank Statement
This demo shows how to quickly load a new bank statement so transactions are automatically recognized and added.

2. Asking About a Specific Expense Type
Ask about your spending in a particular category, such as food, transportation, or utilities.
Example:
User types:
How much did I spend on food last month?

3. Dashboard Notion
You can visual dashboard

Quick Start
1. Clone the repository
git clone https://github.com/your-username/mcp-finance-assistant.git
cd mcp-finance-assistant
2. Install uv (if you don’t have it)
pip install uv
3. Set up the Python virtual environment (client)
uv sync
source .venv/bin/activate # On Windows: .venv\Scripts\activate
4. Install server dependencies and Build
cd servers/finance
npm install
npm run build
cd ../../
5. Configure your environment variables
a. Prepare the .env files
For client/
cp .env-template .env
Then fill out each field in the new .env file:
OPENAI_API_KEY=your_openai_api_key # (optional) For LLM features
MISTRAL_API_KEY=your_mistral_api_key # (optional) For LLM features
NOTION_API_KEY=your_notion_integration_token
NOTION_DB_ACCOUNTS=your_notion_db_accounts_id
NOTION_DB_TRANSACTIONS=your_notion_db_transactions_id
Never commit your
.envfile! Only commit.env-templateas an example.
b. How to get your Notion credentials:
- NOTION_API_KEY:
Create an integration at notion.com/my-integrations and copy your secret. - NOTION_DB_ACCOUNTS and NOTION_DB_TRANSACTIONS:
Copy the database ID from the Notion database URL (everything after your workspace name and before the?).
Make sure you share your databases with the integration you created!
7. Start the client
cd client
uv run main.py
Notion Template
Important:
Before using the assistant, you must duplicate the official Notion template to your workspace.
👉 Click here to duplicate the Finance OS template to your Notion
This project expects two databases in Notion:
- Accounts
- Example properties: Name (title), Type (select), Balance (number), Currency (select), Active (checkbox)
- Transacciones
- Example properties: Name (title), Date (date), Account (relation to Accounts), Amount (number), Type (select), Category (select)
After duplicating the template, make sure to share both databases (“Accounts” and “Transacciones”) with your Notion integration.
Then, copy the IDs of each database and add them to your.envfile as described above.
🤖 Using as an MCP Tool with Claude
MCP Finance Assistant can also be used as an MCP tool, making it available to Anthropic's Claude models for advanced agent workflows.
How to Expose the Finance Tool as an MCP Server
The server component (servers/finance) is fully compatible with the Model Context Protocol (MCP) and can be used as an external tool for Claude or other LLMs supporting MCP.
1. Configure Your .env (as above)
Make sure your Notion API credentials and database IDs are set in the .env for servers/finance.
FAQ
Q: My app won’t connect to Notion!
A: Check your .env credentials, that you shared the databases with your integration, and that the database IDs are correct.
Q: I see “Error during initialization”
A: Make sure the server is running before starting the client, and double-check your dependencies and environment variables.
License
MIT
Author
Jorge Monterroso
Open to contributions and feedback!
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.