BizDesk MCP
Provides structured, read-mostly access to small-business back-office data including customers, invoices, and account notes, allowing Claude to query overdue invoices, revenue summaries, and more.
README
<p align="center"><img src="./assets/banner.svg" alt="mcp-business-server" width="1000"/></p>
<p align="center"> <img src="https://img.shields.io/badge/Python-3.11+-3776AB?logo=python&logoColor=white" alt="Python"/> <img src="https://img.shields.io/badge/MCP-FastMCP-1f6feb" alt="MCP"/> <img src="https://img.shields.io/badge/works_with-Claude_Desktop_·_Claude_Code-D97757?logo=anthropic&logoColor=white" alt="Claude"/> <img src="https://img.shields.io/badge/License-MIT-green" alt="MIT"/> </p>
✅ Verified end-to-end on Windows with local Ollama (
llama3.1,num_ctx=8192) — every command below is from a real run.
BizDesk MCP is a Model Context Protocol server that gives Claude (or any
MCP client) structured, read-mostly access to a small-business back office:
customers, invoices, and account notes. Ask Claude "who owes me money?" and
it calls overdue_invoices() instead of guessing.
MCP is how AI assistants get real hands on real business data — and this repo is the pattern I use for client systems, shrunk to a clean, runnable example.
What it exposes
| Kind | Name | What Claude can do with it |
|---|---|---|
| tool | list_customers(status) |
enumerate active/churned accounts |
| tool | overdue_invoices(min_days_overdue) |
fetch late invoices, most overdue first |
| tool | revenue_summary(year) |
monthly paid revenue + top customer |
| tool | add_note(customer_name, note) |
append a timestamped account note |
| resource | bizdesk://customers/{name} |
full customer card (profile + invoices + notes) |
Design choices worth copying: tools return structured rows, not prose;
the only write path (add_note) is append-only; and the server ships
instructions so the model knows the house rules before it calls anything.
Quickstart
pip install -r requirements.txt
python seed.py # creates bizdesk.db (8 customers, ~50 invoices)
Claude Desktop — add to claude_desktop_config.json (see examples/):
{ "mcpServers": { "bizdesk": { "command": "python", "args": ["/abs/path/server.py"] } } }
Claude Code —
claude mcp add bizdesk -- python /abs/path/server.py
Restart the client, then try: "Which invoices are more than 30 days overdue,
and draft a polite reminder for the worst one?" — watch it chain
overdue_invoices → bizdesk://customers/... → a grounded draft.
Point it at real data
bizdesk.db is plain SQLite. Swap the queries in server.py for your
Postgres/Airtable/QuickBooks layer and the MCP surface stays identical — the
client never knows the backend changed. That's the point of the protocol.
<p align="center">Built by <a href="https://github.com/syedahmad0786">Ahmad Bukhari</a> — AI & Automation Architect · <i>agentic systems that run real businesses, not just demos</i></p>
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。
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。
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。