Brex MCP Server
Enables AI agents to interact with the Brex financial platform, allowing access to account information, transactions, expenses, receipts, budgets, and spend limits through the Brex API.
README
Brex MCP Server
A Model Context Protocol (MCP) server for integrating with the Brex API, enabling AI agents to interact with financial data and resources.
Overview
This MCP server provides a bridge between AI agents and the Brex financial platform, allowing agents to:
- Retrieve account information and transactions
- Access expense data and receipts
- Manage budget resources and spend limits
- View team information
The server implements standardized resource handlers and tools following the MCP specification, enabling secure and efficient access to financial data.
Features
Resources
Account Resources
brex://accounts- List all accountsbrex://accounts/{id}- Access specific account details
Expense Resources
brex://expenses- List all expenses with paginationbrex://expenses/{id}- Access specific expense detailsbrex://expenses/card- List all card expensesbrex://expenses/card/{id}- Access specific card expense details
Note: Expense resources automatically expand merchant and budget information to display human-readable names instead of IDs.
Budget Resources
brex://budgets- List all budgets with paginationbrex://budgets/{id}- Access specific budget detailsbrex://spend_limits- List all spend limitsbrex://spend_limits/{id}- Access specific spend limit detailsbrex://budget_programs- List all budget programsbrex://budget_programs/{id}- Access specific budget program details
Team Resources
brex://users/me- Get current user information
Tools
Receipt Management
match_receipt- Match a receipt with existing expensesupload_receipt- Upload a receipt for a specific expense
Expense Management
update_expense- Update details for a card expense (memo, category, etc.)get_all_expenses- Get all expenses with filtering options and expanded merchant/budget informationget_all_card_expenses- Get all card expenses with filtering options and expanded merchant/budget information
Note: For security reasons, tools that create, update, or delete budgets, spend limits, and budget programs are not implemented in this version.
Installation
Prerequisites
- Node.js v18 or higher
- Brex API access token
Setup
- Clone this repository:
git clone https://github.com/dennisonbertram/brex-mcp-server.git
cd brex-mcp-server
- Install dependencies:
npm install
- Create a
.envfile with your Brex API token:
BREX_API_KEY=your_token_here
BREX_API_URL=https://platform.brexapis.com
PORT=3000
NODE_ENV=development
RATE_LIMIT_REQUESTS=1000
RATE_LIMIT_WINDOW_MS=60000
LOG_LEVEL=info
- Build the server:
npm run build
Configuration with Claude
To use with Claude Desktop, you need to add the server to Claude's configuration file:
On macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
On Windows: %APPDATA%\Claude\claude_desktop_config.json
- Open Claude for Desktop and go to settings by clicking on the Claude menu and selecting "Settings..."
- Click on "Developer" in the left sidebar, then click "Edit Config"
- Update the configuration file with the Brex MCP server settings:
{
"mcpServers": {
"brex-server": {
"command": "node",
"args": [
"/path/to/brex-mcp-server/build/index.js"
],
"env": {
"BREX_API_KEY": "your_brex_api_key_here",
"BREX_API_URL": "https://platform.brexapis.com",
"PORT": "3000",
"NODE_ENV": "development",
"RATE_LIMIT_REQUESTS": "1000",
"RATE_LIMIT_WINDOW_MS": "60000",
"LOG_LEVEL": "info"
}
}
}
}
Make sure to:
- Replace
/path/to/brex-mcp-serverwith the actual path where you installed the server - Replace
your_brex_api_key_herewith your actual Brex API key - Use absolute paths for the server location
Only the BREX_API_KEY and BREX_API_URL values are required; the other environment variables have sensible defaults but can be customized if needed.
- Save the file and restart Claude for Desktop
- Verify the server is working by checking for the hammer icon in the bottom right corner of the input box
Development
For development with auto-rebuild:
npm run dev
Lint your code:
npm run lint
Run tests:
npm run test
Debugging
Since MCP servers communicate over stdio, debugging can be challenging. We recommend using the MCP Inspector for debugging.
Security Considerations
This server implements several security measures:
- Read-only operations for sensitive financial resources
- No storage of API credentials in code
- Rate limiting for API requests
- Proper error handling and logging
Implementation Status
For a detailed implementation plan and status of various features, see documentation/implementation_plan.md.
License
This project is licensed under the MIT License - see the LICENSE file for details.
Author
Dennison Bertram - dennison@dennisonbertram.com
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。
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。
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。
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。
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。
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。
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。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。