
Mcp Builder
- 829 installs
- 7.5k repo stars
- Updated July 31, 2026
- jnmetacode/superpowers-zh
mcp-builder is a skill methodology that teaches developers to design, implement, test, and deploy production-grade Model Context Protocol servers connecting AI assistants to external tools, data sources, and guided workf
About
mcp-builder is a version 1.0.0 MIT-licensed skill that systematizes building Model Context Protocol servers for AI assistants. The guide explains when to expose side-effecting operations as Tools, read-only URI-backed data as Resources, and reusable flows as Prompts, then prescribes a TypeScript layout with src/index.ts entry registration, feature-split tools/ and resources/ folders, shared lib/ clients, and a tests/ directory. Developers reach for mcp-builder when scaffolding a new MCP integration—not when wiring a single REST call—because it covers protocol primitives, structure conventions, validation, and deployment discipline end to end.
- Defines three MCP primitives with clear selection rules: Tools for actions with side effects, Resources for read-only da
- Provides canonical project structures and dependency lists for both TypeScript (@modelcontextprotocol/sdk + zod) and Pyt
- 9-step Tool design checklist covering snake_case verb-first naming, self-describing parameters, enum usage, default valu
- Hard-gate review of tool descriptions before implementation to prevent ambiguous AI selection
- Delivers ready-to-register server entrypoint patterns that immediately work with Claude Code, Cursor and other agent hos
Mcp Builder by the numbers
- 829 all-time installs (skills.sh)
- +29 installs in the week ending Aug 2, 2026 (Skillselion tracking)
- Ranked #1,302 of 16,546 AI & Agent Building skills by installs in the Skillselion catalog
- Data as of Aug 5, 2026 (Skillselion catalog sync)
npx skills add https://github.com/jnmetacode/superpowers-zh --skill mcp-builderAdd your badge
Show developers this skill is listed on Skillselion. Paste this into your README.
| Installs | 829 |
|---|---|
| repo stars | ★ 7.5k |
| Last updated | July 31, 2026 |
| Repository | jnmetacode/superpowers-zh ↗ |
How do you build a production MCP server in TypeScript?
Follow a repeatable methodology for designing, implementing, testing and deploying production-grade Model Context Protocol servers that connect AI assistants to externa
Who is it for?
Backend and platform engineers shipping typed MCP servers that expose external APIs, databases, or workflows to coding agents.
Skip if: Teams only consuming an existing MCP plugin or writing one-off scripts without protocol structure or tests.
When should I use this skill?
The user wants to create, structure, test, or deploy a custom MCP server with Tools, Resources, or Prompts.
What you get
MCP server project structure, registered tools/resources/prompts, test suite, and deployment-ready server package.
- mcp server scaffold
- tool and resource handlers
- test suite
By the numbers
- Defines three MCP primitives: Tools, Resources, Prompts
- Version 1.0.0 MIT license
Files
MCP 服务器构建
系统化设计、实现、测试和部署 Model Context Protocol 服务器的方法论。
1. 协议核心概念
MCP 定义三种原语:
- Tools(工具):AI 助手主动调用的函数,有副作用。如搜索、创建、删除操作。
- Resources(资源):AI 助手只读访问的数据源,用 URI 标识。如
users://{id}/profile。 - Prompts(提示词模板):预定义交互模板,引导用户触发工作流。
选择原则: 执行操作 → Tool | 读取数据 → Resource | 引导交互 → Prompt
2. 项目结构规范
TypeScript
my-mcp-server/
├── src/
│ ├── index.ts # 入口,注册 tools/resources
│ ├── tools/ # 按功能拆分
│ ├── resources/
│ └── lib/ # 客户端封装、校验逻辑
├── tests/
├── package.json
└── tsconfig.json关键依赖:@modelcontextprotocol/sdk + zod
Python
my-mcp-server/
├── src/my_mcp_server/
│ ├── server.py
│ ├── tools/
│ └── lib/
├── tests/
└── pyproject.toml关键依赖:mcp + pydantic
3. Tool 设计原则
命名
snake_case格式,动词开头:search_users、create_issue、delete_file- 名称自解释,AI 助手靠名称选工具,模糊命名导致误调用
参数
- 每个参数有类型约束和
.describe()描述 - 可选参数给默认值,减少 AI 决策负担
- 用枚举代替布尔开关
server.tool("search_issues", {
query: z.string().describe("搜索关键词"),
status: z.enum(["open", "closed", "all"]).default("open").describe("状态筛选"),
limit: z.number().min(1).max(100).default(20).describe("返回上限"),
}, async ({ query, status, limit }) => { /* ... */ });描述
说明用途 + 返回内容 + 限制,这是 AI 选择工具的关键依据:
server.tool("search_users",
"根据姓名或邮箱搜索用户。返回 ID、姓名、邮箱列表。模糊匹配,最多 50 条。",
schema, handler);输出
- 结构化数据 → JSON,人类可读内容 → Markdown
- 始终用
content: [{ type: "text", text: "..." }]格式返回
4. 输入验证和错误处理
用 Zod/Pydantic 做 Schema 级校验,业务级校验放 handler 开头:
server.tool("get_user", { id: z.string() }, async ({ id }) => {
try {
const user = await db.getUser(id);
if (!user) {
return {
content: [{ type: "text", text: `用户 ${id} 不存在,请检查 ID。` }],
isError: true,
};
}
return { content: [{ type: "text", text: JSON.stringify(user, null, 2) }] };
} catch (err) {
return {
content: [{ type: "text", text: `查询失败:${err.message}` }],
isError: true,
};
}
});错误处理四原则: 1. 永远不让服务器崩溃 — try/catch 包裹所有外部调用 2. 返回可操作的错误信息 — 告诉 AI 问题是什么、能做什么 3. 使用 isError: true — 让 AI 知道调用失败 4. 区分错误类型 — 参数错误、权限不足、资源不存在、服务不可用
5. 资源管理和生命周期
// 资源注册
server.resource("user-profile", "users://{userId}/profile", async (uri) => {
const profile = await db.getProfile(extractId(uri));
return { contents: [{ uri: uri.href, mimeType: "application/json", text: JSON.stringify(profile) }] };
});
// 生命周期:先初始化 → 再 connect → 监听关闭信号
const db = await Database.connect(config.dbUrl);
await server.connect(new StdioServerTransport());
process.on("SIGINT", async () => { await db.disconnect(); await server.close(); process.exit(0); });关键点:使用连接池、所有外部调用设超时、优雅关闭清理资源。
6. 测试策略
单元测试 — 业务逻辑与 MCP 注册分离
// tools/search.ts 导出纯函数
export async function searchUsers(query: string, limit: number) { /* ... */ }
// search.test.ts 独立测试
test("返回匹配结果", async () => {
const results = await searchUsers("alice", 10);
expect(results[0].name).toContain("Alice");
});集成测试 — 用 SDK Client 做端到端验证
const [clientTransport, serverTransport] = InMemoryTransport.createLinkedPair();
await server.connect(serverTransport);
const client = new Client({ name: "test", version: "1.0.0" });
await client.connect(clientTransport);
const result = await client.callTool("search_users", { query: "test" });
expect(result.isError).toBeFalsy();MCP Inspector — 交互式调试
npx @modelcontextprotocol/inspector node dist/index.js在浏览器中查看所有 tools/resources,手动调用并查看结果。
测试要点: 每个 Tool 覆盖正常 + 异常路径、边界值、外部服务失败模拟。
7. 安全考虑
权限控制:
- 最小权限原则,读写 Tool 分离
- 危险操作要求确认参数(如
confirm: true)
输入安全:
- SQL 注入 → 参数化查询,绝不拼接
- 路径遍历 → 校验路径,禁止
../ - 命令注入 → 用
execFile而非exec
敏感数据:
- 密钥通过环境变量传入,不硬编码
- 日志不打印完整敏感信息
- 返回数据做脱敏处理
沙箱: 文件操作限制目录、网络请求限制白名单、设置资源配额。
8. 部署和分发
npm 发布
{ "bin": { "mcp-server-myservice": "dist/index.js" }, "files": ["dist"] }用户配置:
{ "mcpServers": { "myservice": { "command": "npx", "args": ["@yourorg/mcp-server-myservice"], "env": { "API_KEY": "xxx" } } } }pip 发布
[project.scripts]
mcp-server-myservice = "my_mcp_server.server:main"Docker — 适用于复杂依赖或隔离场景
FROM node:20-slim
WORKDIR /app
COPY package*.json ./ && RUN npm ci --production
COPY dist ./dist
ENTRYPOINT ["node", "dist/index.js"]9. 调试技巧
关键:MCP 用 stdio 通信,不能用 `console.log`,会破坏协议流。
// 错误
console.log("debug");
// 正确
console.error("[DEBUG]", info);
// 更好
server.sendLoggingMessage({ level: "info", data: "处理中" });常见问题:
| 症状 | 原因 | 解决 |
|---|---|---|
| 启动无响应 | transport 未连接 | 检查 server.connect() |
| Tool 不出现 | 注册在 connect 之后 | 先注册再 connect |
| AI 不调用 Tool | 描述不清晰 | 改善名称和描述 |
| 参数总错 | Schema 不明确 | 添加 .describe() |
| 调用超时 | 外部服务慢 | 加超时和缓存 |
调试流程: Inspector 验证基本功能 → 手动调用确认输入输出 → 连接真实 AI 客户端观察调用模式 → 根据实际行为调整设计。
10. 构建检查清单
设计
- [ ] 明确 Tools vs Resources vs Prompts 分工
- [ ] Tool 命名
动词_名词,描述说明用途和返回内容 - [ ] 参数简洁,可选参数有合理默认值
实现
- [ ] 输入用 Zod/Pydantic 校验
- [ ] 外部调用有 try/catch 和超时
- [ ] 错误返回
isError: true并附可操作信息 - [ ] 不用
console.log(用 stderr 或 SDK 日志) - [ ] 敏感数据走环境变量
测试
- [ ] 核心逻辑有单元测试
- [ ] 有集成测试验证 MCP 协议交互
- [ ] 用 MCP Inspector 手动验证过
- [ ] 用真实 AI 客户端测试过
部署
- [ ] README 含安装和配置说明
- [ ] 提供客户端配置 JSON 示例
- [ ] 遵循 semver,无硬编码密钥
Related skills
How it compares
Use mcp-builder for greenfield MCP server architecture; use a narrower SDK snippet skill when you only need one tool handler patch.
FAQ
When should an MCP capability be a Tool versus a Resource?
mcp-builder assigns side-effecting operations like search, create, or delete to Tools. Read-only URI-identified data such as users://{id}/profile belongs in Resources, while guided multi-step flows use Prompts.
What project layout does mcp-builder recommend?
mcp-builder specifies a TypeScript tree with src/index.ts for registration, src/tools and src/resources split by feature, src/lib for clients and validation, plus tests/ and package.json at the root.