GitHub
GitHub’s official server lets AI read and manage code, issues and pull requests in your account. Use GitHub’s hosted version or run it locally with Docker.
Description from the maker (original English)
Plan and verify multi-ecosystem packaging (Linux, Windows, macOS, HarmonyOS) for AI agents.
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
You need
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Open Claude Desktop → Settings → Developer → Edit Config, and paste this into claude_desktop_config.json
{
"mcpServers": {
"deliverkit": {
"command": "npx",
"args": [
"-y",
"deliverkit-mcp"
]
}
}
}Run this in your terminal
claude mcp add deliverkit -- npx -y deliverkit-mcpPaste into .cursor/mcp.json in your project, or ~/.cursor/mcp.json
{
"mcpServers": {
"deliverkit": {
"command": "npx",
"args": [
"-y",
"deliverkit-mcp"
]
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Paste into .vscode/mcp.json in your project
{
"servers": {
"deliverkit": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"deliverkit-mcp"
]
}
}
}Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This one runs on your computer, and ChatGPT can only connect to servers on the internet. Pick another app.
Your choice stays on this device in a cookie; no account needed.
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
Runs on your computer; the tool list appears in your AI app after you install it.
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 8, 2026
The registry name was verified with the GitHub account muzimu217, which owns the repo
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
This one runs on your computer; there is no internet endpoint
Runs on your computer and uses local settings instead
Runs on your computer; we can only read tool lists from internet endpoints
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 1 of 1 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
The registry does not list a required key, though some servers ask you to sign in on first use.
It runs on your computer. Your AI app starts it when needed, and data does not pass through the maker’s servers (unless it calls an online service itself).
The GitHub repo last changed –, and the latest version (0.3.2) reached the registry 2 months ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need Node.js on your computer.
In the same category: GitHub, Desktop Commander, Cloudflare MCP
Servers in the same category from other makers, sorted by stars.
Reading a card:↓ Downloads per weeki★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
GitHub’s official server lets AI read and manage code, issues and pull requests in your account. Use GitHub’s hosted version or run it locally with Docker.
MCP server for terminal commands, file operations, and process management
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
AI 交付大脑:给 Agent 用的全生态交付编排工具。
DeliverKit 不是一个"把 Linux 软件塞进 Windows 跑"的兼容层,而是一个给 AI(Agent)使用的交付大脑:它把各生态(Linux / Windows / 苹果 / 鸿蒙)的打包、签名、上架知识变成 AI 能读懂、能执行、能验证的能力,让 AI 理解用户需求后,规划出一条合法合规的交付链路,把已经开发好的产品送达每一个生态。
徽章实时取自 npm 注册表;下列条目均带可点验链接。项目无默认遥测,所有采用数据都来自公开渠道(npm / GitHub / 目录收录),每周快照存档于 metrics/。
npx -y deliverkit-mcp 即可验证安装真实可用)verified 状态的 ReleaseManifest 证据(artifact 可下载)发布于 2026-09-11。上面的徽章与链接就是当前采用状态的全部事实——我们只展示可验证的数据,不刷不夸。
npx -y --package=deliverkit-mcp -- deliverkit doctor # 先看本机现在能交付哪些目标,缺什么、怎么补
npx -y --package=deliverkit-mcp -- deliverkit inspect . # 识别项目语言与入口 npx -y --package=deliverkit-mcp -- deliverkit plan . --goals deb,rpm # 生成可评审的 Forge.md 交付契约 npx -y --package=deliverkit-mcp -- deliverkit pack-deb . --plan Forge.md # 构建 + 干净容器安装运行验证(需 Docker)
doctor 会把「这台机器现在能产出哪些包」摊开讲清楚——不能产出的目标不是缺陷,而是各生态用签名和官方工具链锁定了产出位置,DeliverKit 的做法是把它们规划到正确的 runner 上,而不是在本机伪造产物。
失败时不会只丢一个日志路径给你:错误结果里带 log_excerpt(日志尾部片段)、suggested_fix 和 next_actions,Agent 和人都能直接看到原因与下一步。
"一次开发、各生态安装包自动产出"之所以难,不是因为没人想做,而是各生态官方用代码签名 + 官方工具链 + 官方账号把这条路锁死了:
| 生态 | 产出安装包的硬约束 |
|---|---|
| 苹果(macOS/iOS) | 签名与公证只能在 macOS + Xcode + Apple 开发者账号上完成 |
| Windows | 代码签名需要 Windows 代码签名证书 |
| 鸿蒙 | 需要 华为 DevEco 工具链 + AGC 签名 |
| Linux(deb/rpm/AppImage) | 唯一能在一台机器上多目标产出,无强制签名门槛 |
所以 DeliverKit 不假装一台机器产出全生态——它懂得"每个安装包该在哪、用什么合法方式产出、怎么验证它真的能装能跑",并指挥 AI 在正确的环境(本地 / CI 对应平台 runner / 云构建)里完成。这就绕开了各生态的硬约束,而不是去翻墙。
DeliverKit 通过 MCP(stdio)暴露规划、编排与构建工具:
language(python/javascript/typescript/go/arkts)与 entrypoints 手动指定,结果直接进入 Forge.md 契约,不再死路。Forge.md 交付契约(目标生态 / 产物 / 决策依据 / 风险)。Forge.md 在隔离 Ubuntu 容器中构建 .deb,并在新容器执行安装和运行验证。支持 systemd 参数:同时安装 systemd 服务单元与维护脚本(真实主机上 postinst 自动 enable --now),验证阶段含 systemd-analyze 语法校验。Forge.md 在隔离 Rocky Linux 容器中构建 .rpm,并在新容器执行安装和运行验证。支持 systemd 参数:unit 进 %files(/usr/lib/systemd/system),%post 在真实主机自动 enable --now,验证阶段含 systemd-analyze 语法校验。Forge.md 在隔离 Ubuntu x86_64 容器中构建 AppImage,并执行 extract-and-run/AppRun 验证。Forge.md 生成 GitHub Actions 多平台工作流;Windows/Apple 证书只通过 secrets 注入,HarmonyOS job 要求 Linux self-hosted DevEco runner。ReleaseManifest.json。三个 Linux 构建工具都需要可用 Docker 守护进程。npm run test:e2e:linux 已用 Python Flask、TypeScript 和 Go 标准库 HTTP 三个真实 fixture 复现同一份多目标 Forge.md,完成 deb/rpm/AppImage 的构建、安装与运行验证,并生成 verified 状态的 Release Manifest;仓库 CI 也会在 Ubuntu 22.04 与 24.04 runner 上重复该 matrix 并上传 JSON 证据,其他 Linux 主机复核仍建议继续执行。
所有平台构建类工具都遵守两条铁律:
Forge.md 契约不构建。失败时的可行动性也被当作契约的一部分:Docker 不可用会区分「没装 / 守护进程没起 / 权限不足 / 探测超时」并给出各自的修复动作;构建超时不会伪装成普通构建失败;每次失败都带日志尾部片段与下一步命令;Forge.md 的 source_dir 比对容忍符号链接差异(如 macOS 的 /tmp 与 /private/tmp),真不一致时报错会列出两个路径各自的值。
CLI 侧同样支持手动指定(可重复 --entry):
npx -y --package=deliverkit-mcp -- deliverkit plan . --goals deb --language python --entry server.rb
仓库附带可直接部署的静态站点:site/index.html。首屏是 MCP 安装命令与「10 分钟跑通 Linux」首次任务,生态知识与平台前置指南作为次级入口,Agent Skill 见 skills/deliverkit-mcp/SKILL.md。
npx serve site
站点不收集任何用户数据(无表单、无统计脚本、无 cookie),也不包含任何账号、证书或私钥;部署与内容边界见 site/README.md。
MCP(stdio):
{
"mcpServers": { "deliverkit": { "command": "npx", "args": ["-y", "deliverkit-mcp"] } } }
CLI:
deliverkit doctor # 自检本机可交付目标,缺什么、怎么补
deliverkit inspect . # 识别项目与交付目标建议 deliverkit plan . --goals deb,rpm # 生成 Forge.md 交付契约 deliverkit pack-deb . --plan Forge.md # 构建并验证 deb(需要 Docker) deliverkit pack-rpm . --plan Forge.md # 构建并验证 rpm(需要 Docker) deliverkit pack-appimage . --plan Forge.md # 构建并验证 AppImage(需要 Docker) deliverkit generate-ci-workflow . --plan Forge.md # 生成多平台 GitHub Actions 工作流 deliverkit pack-windows-msi . --plan Forge.md # 仅 Windows runner 可执行,要求签名 secrets deliverkit pack-macos . --plan Forge.md # 仅 macOS runner 可执行,按 Forge.md 构建 DMG/PKG deliverkit pack-harmonyos . --plan Forge.md # 仅 DevEco runner 可执行,要求 AGC secrets/设备 deliverkit generate-release-manifest . --plan Forge.md --results .deliverkit/results
CLI 默认输出人类可读摘要(产物、验证项、失败原因、下一步);任何命令加 --json 得到与 MCP 一致的结构化结果,供脚本和 Agent 消费。
见 docs/ROADMAP.md。简要:
DeliverKit 的 MCP 协议层与"计划先行 / 真实验证"两条铁律继承自 ForgeKit(一个专注 Docker 构建诊断的项目)。两者定位不同、各自独立演进:ForgeKit 往窄里做诊断,DeliverKit 往宽里做多生态交付。
MIT