expense-tracker
A lightweight MCP server for recording, browsing, and summarizing expenses locally using SQLite, with a customizable categories resource.
README
Expense Tracker MCP Server
A lightweight Model Context Protocol (MCP) server for recording, browsing, and summarizing expenses. It stores data locally in SQLite and exposes a curated category list as an MCP resource, so an MCP client can help keep spending records organized without relying on a hosted service.
Features
- Add expenses with a date, amount, category, optional subcategory, and note
- List expenses within an inclusive date range
- Summarize spending by category, with an optional category filter
- Read the available categories and subcategories from
expense://categories - Persist everything locally in
expenses.db
Requirements
- Python 3.13 or newer
uv(recommended)
Install
Clone the repository and install the project dependencies:
git clone https://github.com/ansh-rohilla/expense-tracker-mcp-server.git
cd expense-tracker-mcp-server
uv sync
Run the server
Start the server with:
uv run python main.py
The SQLite database is created automatically at expenses.db in the project root on first run.
Configure an MCP client
Add the following server entry to your MCP client's configuration. Replace /absolute/path/to with the path where you cloned this repository.
{
"mcpServers": {
"expense-tracker": {
"command": "uv",
"args": [
"--directory",
"/absolute/path/to/expense-tracker-mcp-server",
"run",
"python",
"main.py"
]
}
}
}
For example, in Claude Desktop this entry belongs in claude_desktop_config.json. Restart the client after saving the configuration.
Available tools
add_expense
Adds an expense to the local database.
| Parameter | Required | Description |
|---|---|---|
date |
Yes | Date of the expense, preferably YYYY-MM-DD |
amount |
Yes | Expense amount as a number |
category |
Yes | A top-level category such as food or transport |
subcategory |
No | A more specific category, such as groceries |
note |
No | Any useful context about the expense |
Example request:
Add an expense of 450 on 2026-08-02 for groceries under food. Note: weekly vegetables.
list_expenses
Returns expenses whose dates fall within the inclusive range.
| Parameter | Required | Description |
|---|---|---|
start_date |
Yes | Beginning of the date range (YYYY-MM-DD) |
end_date |
Yes | End of the date range (YYYY-MM-DD) |
summarize
Returns total spending grouped by category for an inclusive date range.
| Parameter | Required | Description |
|---|---|---|
start_date |
Yes | Beginning of the date range (YYYY-MM-DD) |
end_date |
Yes | End of the date range (YYYY-MM-DD) |
category |
No | Limit the summary to one category |
Categories resource
The server provides expense://categories as JSON. It reads from categories.json, which you can customize to match your own budgeting system. Common top-level categories include food, transport, housing, utilities, health, education, entertainment, shopping, travel, investments, and more.
Data and privacy
Expenses remain on your machine in the local SQLite database. Back up expenses.db if you want to retain your records when moving or reinstalling the project. Because the database can contain personal financial information, avoid committing a populated copy to a public repository.
Project layout
.
├── main.py # FastMCP server and expense tools
├── categories.json # Available categories and subcategories
├── expenses.db # Local SQLite database (created automatically)
├── pyproject.toml # Project metadata and dependencies
└── uv.lock # Locked dependency versions
License
No license has been specified yet. Add a license file before distributing or reusing this project publicly.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。