carbon-footprint-mcp
An MCP server for calculating carbon footprints from bank statements using EPA GHG emission factors.
README
Carbon Footprint Calculator (MCP Server)
An MCP (Model Context Protocol) server for calculating organizational carbon footprints from bank statements, financial exports, and structured activity data using EPA GHG emission factors.
Privacy and security first
- Runs 100% locally on your machine or server
- Sends no financial data to external APIs or cloud providers
- Stores no data by default
- Exposes read-only calculation and reporting tools
- Works with Claude Desktop, Cursor, and other MCP clients
<!-- mcp-name: io.github.MayankTalwar0/carbon-footprint-mcp -->
Why This Exists
If you are preparing ESG reporting, investor diligence materials, or internal sustainability reviews, getting to a usable emissions baseline is usually slow and manual.
This server helps turn raw bank statements, Xero or QBO exports, and structured operational inputs into a carbon footprint report in minutes. It maps activities to EPA-aligned emission factors and produces both HTML and Markdown outputs.
The user experience is designed to work for organizations in any country, while the current electricity benchmarking still uses EPA eGRID regional factors under the hood.
What It Does
- Ingests bank CSVs, Xero or QBO exports, and structured activity data.
- Helps categorize transactions into likely emission sources such as electricity, fuel, travel, shipping, and waste.
- Computes Scope 1, Scope 2, and Scope 3 emissions using EPA GHG emission factors.
- Scores carbon intensity by revenue and headcount when those inputs are provided.
- Generates polished HTML and Markdown reports.
Emission Factor Source
All emission factors are based on the EPA GHG Emission Factors Hub (January 2025), including eGRID 2023 electricity factors and IPCC AR5 global warming potentials.
Covered categories include stationary combustion, mobile combustion, electricity, steam or heat, transportation, waste disposal, business travel, employee commuting, and refrigerants.
Installation
Claude Desktop
- Install
uv. - Open Claude Desktop settings and edit the MCP config.
- Add this server:
{
"mcpServers": {
"carbon-footprint": {
"command": "uvx",
"args": ["carbon-footprint-mcp"]
}
}
}
- Restart Claude Desktop.
Claude Code or Cursor
claude mcp add carbon-footprint -- uvx carbon-footprint-mcp
Local Development
git clone https://github.com/MayankTalwar0/carbon-footprint-mcp.git
cd carbon-footprint-mcp
pip install -e .
carbon-footprint-mcp
Available MCP Tools
| Tool | Description |
|---|---|
computeEmissions(inputs_json) |
Computes GHG emissions from structured activity data across all 3 scopes. |
generateEmissionsReport(emissions_json, output_dir) |
Renders a polished HTML and Markdown report and saves it to disk. |
listEmissionFactors(category) |
Lists available fuel, eGRID, and waste emission factors. |
Supported Emission Categories
| Scope | Category | Input Required |
|---|---|---|
| 1 | Stationary Combustion | Fuel type and quantity |
| 1 | Mobile Combustion | Fuel type and gallons |
| 1 | Refrigerant Leakage | Gas type, leaked kg, and GWP |
| 2 | Purchased Electricity | kWh and eGRID subregion |
| 2 | Purchased Steam or Heat | mmBtu |
| 3 | Transportation and Distribution | Vehicle type and distance |
| 3 | Waste Disposal | Material, short tons, and disposal method |
| 3 | Business Travel | Travel mode and passenger-miles |
| 3 | Employee Commuting | Commute mode and passenger-miles |
Carbon Intensity Scoring
| Score | tCO2e per $1M Revenue | Interpretation |
|---|---|---|
| Excellent | < 5 | Best-in-class for low-footprint operations |
| Good | 5-20 | Low intensity |
| Moderate | 20-100 | Typical for services and tech |
| High | 100-500 | Heavy operations |
| Very High | > 500 | Very high intensity |
License
MIT
Built By SlickBooks
Built by Mayank, founder of SlickBooks.
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。