mcp-gnu-units
Unit converter for AI agents, powered by the GNU units database — convert 3000+ units of measurement, evaluate compound unit expressions, reduce to SI base units, dimensional analysis. Offline, deterministic.
README
mcp-gnu-units
mcp-gnu-units is an MCP server that gives AI agents precise, offline unit
conversion backed by the GNU units database. Convert between 3000+ units of
measurement — length, mass, time, temperature, area, volume, energy, power,
pressure, speed, data and more — evaluate compound unit expressions like kW*h
or acre*ft, reduce any quantity to its SI base units, search the unit database
by keyword, and look up the exact definition of any unit, prefix, or physical
constant. Deterministic, dimension-aware, and free of the hardcoded per-category
unit tables every other converter ships.
Status: feature-complete, pre-release. All five domain tools ship —
find_units,convert,convert_to_si,define_unit, andlist_prefixes— alongside theinfohealth check. Not yet published to PyPI; the remaining publish steps are tracked inTODO.
Distribution name: mcp-gnu-units · import package: mcp_gnu_units.
Built on GNU units
This project stands on the shoulders of GNU units, the units-conversion program written and maintained by Adrian Mariano for the Free Software Foundation. Its database is the product of decades of careful curation — thousands of units, physical constants drawn from CODATA and NIST, and the conversion semantics that make dimensional analysis trustworthy. This server would not exist without it.
mcp-gnu-units bundles that database verbatim (see
Bundled data & provenance below) and pairs it with
an independent, pure-Python conversion engine, so every number it returns traces
straight back to GNU units' own definitions. It is offered in the same spirit
and under the same license (GPL-3.0-or-later). If you find this useful, the
original earns the credit first: https://www.gnu.org/software/units/.
Bundled data & provenance
This server stands on GNU units, Adrian
Mariano's units-conversion program, by bundling its unit database verbatim at
src/mcp_gnu_units/data/definitions.units. The conversion engine here is an
independent implementation; GNU units supplies the data.
- Source: GNU units 2.27 (tarball),
definitions.unitsdata version 3.26 (2026-02-25). - License: GPL-3.0-or-later © Free Software Foundation — the same license as this project, so the two are wholly compatible.
- Integrity: the upstream tarball was GPG-verified (good signature from Adrian Mariano) before the file was vendored.
Full attribution, version pins, and checksums are recorded in
NOTICE.
Sponsoring
mcp-gnu-units is free, open-source software developed in my spare time. Sponsoring keeps this project alive and actively maintained — it funds new units-engine features, bug fixes, and ongoing support, and it's a direct signal that the work is worth continuing.
If the project is useful to you, please consider sponsoring it through GitHub Sponsors. Click the Sponsor button at the top of the repository, or visit the link directly, and pick a one-time or recurring tier. Every contribution, large or small, is hugely appreciated and goes straight back into keeping mcp-gnu-units healthy.
License
GPL-3.0-or-later. See LICENSE. The bundled GNU units database is
also GPL-3.0-or-later © Free Software Foundation; see NOTICE.
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。