经 AI Skill Hub 精选评估,MCP开发框架 获评「推荐使用」。这款MCP工具在功能完整性、社区活跃度和易用性方面表现出色,AI 评分 7.5 分,适合有一定技术背景的用户使用。
MCP开发框架 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
MCP开发框架 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/raiworks/mastery.md
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"mcp----": {
"command": "npx",
"args": ["-y", "mastery.md"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 MCP开发框架 执行以下任务... Claude: [自动调用 MCP开发框架 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"mcp____": {
"command": "npx",
"args": ["-y", "mastery.md"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
Current: v3.6.0 · Changelog · Website
A universal, tech-agnostic development process framework — one Markdown file that defines how software projects should be structured, discussed, designed, planned, built, shipped, and reflected upon.
mastery.md and mastery-compact.md into your project's docs/ directorydocs/project-discussion.md using the embedded templateyour-project/
├── docs/
│ ├── mastery.md # Full framework (read-only — never edit this)
│ ├── mastery-compact.md # Compact variant for AI agents (read-only)
│ ├── project-discussion.md # First: discuss what you're building
│ ├── project-context.md # Then: formalize the project identity
│ ├── project-roadmap.md # Then: plan the feature order
│ └── features/ # Per-feature lifecycle docs
│ └── 01-feature-name/
├── src/ # Your code
└── ...
| Aspect | Without Mastery | With Mastery |
|---|---|---|
| **Project start** | Jump straight into code; goals live in someone's head | Structured discussion doc captures the why, what, and constraints before a line is written |
| **Context between sessions** | AI agents (and humans) lose track — repeat questions, duplicate work | Docs act as persistent memory; anyone picks up exactly where work stopped |
| **Feature planning** | Ad-hoc tickets or mental checklists | Each feature gets a lifecycle folder: discussion → architecture → tasks → test plan → changelog |
| **AI agent behavior** | Agents guess at conventions, hallucinate structure, skip steps | Agent Protocol defines what to read, what to decide, and when to ask for help |
| **Git workflow** | Inconsistent branch names, vague commits | Defined branching strategy and commit format baked into the process |
| **Design decisions** | Lost in chat threads or never recorded | Architecture docs with explicit rationale — searchable, reviewable, permanent |
| **Onboarding** | "Read the code" or a stale wiki | New contributors read the docs folder and understand the full project state in minutes |
| **Session handoff** | "Where were we?" every time | Changelog session notes and task checkboxes show exactly what happened and what's next |
| **Process consistency** | Every project reinvents its workflow | One file, same lifecycle, any tech stack |
| **Tooling required** | CI configs, project management apps, custom templates | Pure Markdown — zero dependencies, works everywhere |
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
AI Skill Hub 点评:MCP开发框架 的核心功能完整,质量良好。对于Claude Desktop / Claude Code 用户来说,这是一个值得纳入个人工具库的选择。建议先在非生产环境试用,再逐步推广。
| 原始名称 | mastery-md |
| 原始描述 | 开源MCP工具:A universal, tech-agnostic development process framework distributed as a single。⭐6 |
| Topics | mcpai-agentscopilotcursordeveloper-toolsproject-management |
| GitHub | https://github.com/raiworks/mastery.md |
| License | MIT |
收录时间:2026-05-27 · 更新时间:2026-05-30 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端