Skip to main content
Glama
Ajiumeng123

dragon-work

by Ajiumeng123
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Local

Dragon Work

本地 AI 工作台：对话、Skill、MCP 与项目管理。桌面壳基于 Tauri + Vue，核心能力基于 Java 生态，内核使用阿里巴巴 AgentScope 开发。

数据写在本机用户目录 .dragon_work，不经过云端。

内核仓库：dragon-work-core

主题素材声明

本仓库中的主题、立绘、图标等视觉素材由 AI 生成。若权利人认为构成侵权，请联系仓库维护者，我们会及时删除相关素材。

Related MCP server: Helix MCP Server

仓库结构

路径

说明

src/

桌面前端（Vue）

src-tauri/

Tauri / Rust 壳，托盘、打包、拉起内核

src-tauri/resources/backend/

内核 JAR 放置目录（不提交 git）

src-tauri/resources/jre/

内嵌 JRE 放置目录（不提交 git）

内核打包后放哪里

内核在 dragon-work-core 构建。打出 Spring Boot fat jar 后，复制到本仓库：

src-tauri/resources/backend/dragon-work-start-0.0.1.jar

要求：

  • 目录固定为 src-tauri/resources/backend/

  • 文件名以 .jar 结尾即可（启动时会取该目录下第一个 jar）

  • 不要提交 jar，已写入 .gitignore

同时放入完整 JRE（需包含 conf/security/java.security），例如 Temurin 21：

src-tauri/resources/jre/
├── bin/java.exe
├── bin/javaw.exe
├── conf/security/java.security
└── lib/

npm run tauri build 会把 backend/ 与 jre/ 打进安装包。安装后程序会把运行时镜像到 %LOCALAPPDATA%\com.wenzixiu.dragon-work\embedded-runtime\，避免安装路径含中文导致 JVM 无法加载 java.security。

开发与打包

开发模式 不启动 jar，只连接本机已运行的内核：http://127.0.0.1:17623/dragon-work

# 先启动 dragon-work-core
npm install
npm run tauri dev

强制开发时也拉起内嵌 jar：

set DRAGON_EMBED_BACKEND=1
npm run tauri dev

打包安装包（会内嵌 jar + JRE）：

# 确认 backend/*.jar 与 jre/ 已就位
npm run tauri build

内核默认端口 17623，context-path /dragon-work。

环境

  • Node.js 18+

  • Rust（Tauri）

  • 开发：本机 Java 21 跑内核

  • 打包：把 JRE 21 整目录放到 src-tauri/resources/jre/

效果

This server cannot be deployed

Pay for results, not tokens.Fund a bounty. Let AI agents compete and collaborate to solve your problems.agentbounties.appAd

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • BridgeDesk

    stanic-mah
    A
    license
    Not graded
    quality
    A
    maintenance
    A desktop launcher and local MCP workspace server that enables ChatGPT and other MCP clients to securely read, edit, search, run commands, and show changes in selected local project folders.
    1
    MIT

  • Helix MCP Server

    b1krams
    F
    license
    A
    quality
    D
    maintenance
    A local-first MCP server providing secure workspace file operations, offline full-text search, and web search/fetch capabilities without requiring API keys.
    10
    -
  • sthayi by sthayi-ai

    sthayiofficial

    sthayi-ai
    A
    license
    Not graded
    quality
    A
    maintenance
    A local-first MCP server that gives AI clients a shared, user-owned memory and skills plane, enabling persistent memory across apps via the Model Context Protocol.
    MIT

  • CodeLocal

    skyrocketeer
    A
    license
    Not graded
    quality
    B
    maintenance
    Connects MCP-compatible AI clients and agents to one durable Project Brain plus controlled access to authorized project folders on your own machine. Enables file editing, terminal and Git operations, project intelligence, browser automation, and local MCP server federation, all scoped to explicit workspace grants.
    56 npm
    Apache 2.0