graftcode

MCP其他

提供方 grft-dev · 包名 graftcode

其他 stdio · 本地进程 未标注 社区 / 第三方 Claude DesktopCursorClineClaude Code
查看接入方式

Connect any code. Call as dependency. The official repository for Graftcode: the integration-free development layer that lets you call methods across languages and runtimes on remote services as if they were local. Eliminate integration lay

一句话选型结论

按资料字段整理

graftcode 属于其他类 MCP 服务器,stdio · 本地进程,鉴权方式未标注,可直接接入 Claude Desktop、Cursor、Cline 等,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类其他 传输方式stdio · 本地进程 鉴权未标注 部署方式本地部署
GitHub 星标 205 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

其他
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

ai-agents developer-experience devtools graftcode integration

支持客户端

适用场景

ai-agentsdeveloper-experiencedevtools

安装与配置

stdio · 本地进程 未标注
客户端在本机拉起子进程,数据不出本机
git clone https://github.com/grft-dev/graftcode && cd graftcode

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 Cross-platform,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
接入外部数据

把外部系统的数据接进对话。

自动化重复操作

把固定步骤交给模型跑。

补充工具能力

给模型补上它原本没有的能力。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Cursor、Cline,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在其他场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Cursor 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
graftcode当前 未标注 stdio · 本地进程 本地部署 有 2026-08-23
gemini-cli 未标注 stdio · 本地进程 本地部署 有 2026-04-13
ext-apps 未标注 stdio · 本地进程 本地部署 有 2026-04-19
northcinder 未标注 stdio · 本地进程 本地部署 有 2026-08-23
fastctx 未标注 stdio · 本地进程 本地部署 有 2026-08-23
quickstart-resources 未标注 stdio · 本地进程 本地部署 有 2026-04-12

服务器档案

其他
字段来自项目仓库与公开索引,中文说明由本站标注
提供方grft-dev上游标注的维护方 grft-dev
分类其他本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式未标注远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时graftcode · Cross-platform本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-08-23上游最后核验时间

详细介绍

7.4 KB
来自项目仓库与上游目录的原文,未做删改

graftcode


Call remote code like local dependencies.

Graftcode is the integration-layer-free development framework that lets you call methods across any language and runtime as if they were local, without writing integration code, SDKs, or client layers.

Supported Languages

Platforms & OS

Cloud Ready

Call remote code like local dependencies. Let package managers keep them fresh.

Graftcode is the integration-free development layer that lets you call methods across languages and runtimes as native dependencies: no integration layer code, no SDKs, no client code.

Turn your public methods into instantly consumable, strongly-typed clients in any technology using regular package managers.

Access your remote services from web, mobile, and edge clients. Seamlessly connect microservices and expose business logic the way it was always meant to be clean, type-safe, and effortless.

Imagine a world where you can share business logic across any technology as easily as publishing a package to a public repository. With Graftcode, you simply write public methods, communicate errors through exceptions, and call them like local code.

Graftcode removes the need to hand-write REST, gRPC, or Thrift integration layers, or to couple your code to message buses and event systems. It reduces your codebase by up to 50%, dramatically improves AI-assisted development (lower token usage, higher efficiency, and simpler code reviews), and makes every backend immediately MCP-compatible for AI agents.

It also enables true modular monoliths you can decompose and recombine modules at will, switch communication channels (WebSocket, HTTP/2, RabbitMQ, Kafka, Service Bus, SQS, Pub/Sub, etc.) without changing a single line of code, and build truly polyglot systems with zero integration overhead.


What is Graftcode?

Instead of building and maintaining APIs, DTOs, clients, and integration layers, Graftcode lets you:

  • Write public methods in your preferred language
  • Run them on Graftcode Gateway (multi-runtime host running on your cloud instances, bare metal, containers or local machine)
  • Install auto-generated, strongly-typed Grafts in any other project via npm, dotnet, pip, etc.
  • Call remote code as if it was local

No controllers. No OpenAPI. No client code. No IDL.

Graftcode works with monoliths, microservices, and everything in between — and comes with built-in MCP support for AI agents.


Quick Start

1. Host your service```bash

1. Build/package your library

2. Navigate to the output folder

3. Download Graftcode Gateway

Windows

iwr https://grft.dev/get | iex

Linux / macOS

curl -fsSL https://grft.dev/get | sh

Run the Gateway with your module:

Example with .NET

gg --modules ./MyService.dll


### 2. Consume it from anywhere
Install a Graft using your package manager (example for npm):```bash
npm install --registry https://grft.dev/YOUR_PROJECT_ID @graft/your-service-name

Graft installed command can be obtained by navigating to your service hosted on Graftcode Gateway i.e. http://localhost if executed locally.

3. Call it like local code

Once you install Graft in client project of your choice, just import classes replicating your remote API and call the methods as if they where available as local module.```typescript
import { EnergyPriceCalculator } from '@graft/...';

const price = await EnergyPriceCalculator.GetPrice();


Whenever API evolves, graftcode will take care to inform your package manager that new version of dependency is available.

### Interactive Quick-Start Journey

The best way to get started with Graftcode is our hands-on tutorial series.

**→ Open Quick Start Guide**

It contains step-by-step tutorials such as:
- Connecting frontend to backend
- Exposing your first service
- Switching between monolith and microservices
- Using with AI agents (MCP)

## Core Repositories

| Repository                        | Purpose                                      |
|-----------------------------------|----------------------------------------------|
| graftcode-gateway | Core multi-runtime host (the engine)        |
| graftcode-demos     | Live demos + performance comparisons        |
| graftcode-quick-start-guide | Hands-on tutorials |
| graftcode-documentation | Full documentation |
| graftcode-docker-images | Official Docker images |
| graftcode-plugins | Official Open Source plugins for any communication channel (ServiceBus, RabbitMQ, Kafka etc..) |

## Key Features

- Cross-language & cross-runtime method calls
- Auto-generated, always-in-sync strongly-typed clients
- Monolith ↔ Microservices switching with zero code changes
- Built-in **MCP** support for AI agents
- Up to 70% faster & significantly lower CPU usage than REST/gRPC
- Works with 20+ languages and 10+ package managers
- 50% reduction in codebase volume
- 30-70% lower cost of tokens for any change lead by AI
- significantly higher success rate of AI-assisted development, leveraging better context efficiency

## Exposing your methods for AI Agents (MCP)

Graftcode has native support for **Model Context Protocol (MCP)**. Any public static methods with value type arguments and results become callable as MCP tools. Run your module on Graftcode Gateway, enter the service address and copy MCP configuration from Graftcode Vision portal exposed by Graftcode Gateway.

### Ready to use AI Agent rules

Adding graftcode rules to your repository allows you to build entire monolith or distributed systems based on graftcode using pure prompts — the AI agent reads the rules and can easily build new systems in greenfield project, migrate existing solutions to graftcode, and evolve your solution with significantly higher context efficiency and lower token usage.

Use command below in your project root directly to download rules for any IDE Agent.```bash
# PowerShell
iwr grft.dev/get | iex

# Linux / macOS
curl -fsSL grft.dev/get | sh

# Alternative with wget
wget -qO- grft.dev/get | sh

AI Agent Rules Raw Files

If you want to access the raw files of our ready-to-use rules and instructions for AI coding agents that are getting installed when you call abovementioned one-liner command, check the folder below:

→ /rules

These files help AI agents (Cursor, Claude, etc.) understand and effectively use Graftcode when generating or refactoring code.


Links

  • Website: https://graftcode.com
  • Portal: https://portal.graftcode.com
  • Academy / Quick Start: https://academy.graftcode.com
  • Docs: https://docs.graftcode.com
  • Discord: Join Community
  • X / Twitter: @Graftcode
  • LinkedIn: Graftcode
  • Public Roadmap: https://graftcode.featurebase.app/dashboard/roadmap

Used by

Graftcode is already trusted by innovative teams at:

(and many more engineering teams building the future of software)


⭐ Star us if you like the project

If Graftcode helps you ship faster and cleaner or just inspire you with its bold vision — please star the repository.
It really helps the project grow and motivates the team.

Contributing

We welcome contributions! Please read our Contributing Guide and feel free to open issues or PRs.

Made with ❤️ for developers who want to focus on business logic instead of integration layers.

Join the Community

Join other Grafters building the future of software integration.

Chat with the team, get help, share ideas, and stay up to date with the latest developments.

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
graftcode 怎么安装?

在客户端按下面的安装命令接入:git clone https://github.com/grft-dev/graftcode && cd graftcode。对应的包名是 graftcode。装完在客户端配置里引用同一个包或端点。

graftcode 怎么在 Claude Desktop、Cursor、Cline 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

graftcode 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

graftcode 需要 API Key 吗?

资料没有标注鉴权方式,是否需要密钥取决于服务本身,接入前按官方文档确认。

graftcode 支持哪些客户端?

资料标注支持 Claude Desktop、Cursor、Cline、Claude Code;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

graftcode 能提供哪些能力?

资料显示能力标签包括 ai-agents、developer-experience、devtools、graftcode、integration。具体工具清单以官方文档为准。

graftcode 需要什么运行环境?

运行时为 Cross-platform;属于本地进程,需要在本机装好运行环境。

graftcode 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-08-23。