@ 后端kit- labs/ mcp- 服务器

MCP(模式上下文协议)服务器将代码智能,技能管理和项目上下文暴露为任何MCP相容的AI客户端(Claude Desktop,代理-core,Cursor等)的工具.

状态 : stable ——保持并做好生产准备. 在试验阶段没有新的特点.

目录

• 安装
• 运行服务器
• 工具参考
• 代码情报索引
• 通过代理核心连接
• 通过 Claude 桌面连接
• HTTP 模式
• 环境变量

安装

npm install -g @backendkit-labs/mcp-server
# or use via npx:
npx @backendkit-labs/mcp-server

二进制 : bk-mcp

运行服务器

Stdio 模式( 默认—— 用于 Claude 桌面, 代理核心)

bk-mcp

# Against a specific project
BK_CWD=/my-project bk-mcp

# With a custom app name
BK_APP_NAME=my-app BK_CWD=/my-project bk-mcp

HTTP 模式( 多会话)

bk-mcp --http --port 3100

工具参考

stack_detect

使用文件检测和代码索引检测项目的技术堆栈.

Args:
  cwd  string?  Project directory (defaults to server cwd)

示例

What stack does this project use?
→ stack_detect()
→ "Detected stack: typescript, node, react, docker"

code_index

将代码库索引为 SQLite 图表数据库。 递增式——只有再索引更改了文件.

Args:
  cwd    string?   Project directory
  watch  boolean?  Enable file watcher for live updates (default: false)

示例

Index this codebase
→ code_index({ watch: true })
→ "Indexed 234 files, 4,821 symbols. File watcher enabled."

被索引的内容 : 函数,类,接口,类型,来自TypeScript/JS/Go/Python/Java/Kotlin的enum. 每个符号(呼叫图)都记录有呼叫者和被呼叫者.

code_search

以名称或关键词搜索符号 .

Args:
  query  string   Symbol name or keyword
  limit  number?  Max results (default 20)

示例

Where is TokenValidator defined?
→ code_search({ query: "TokenValidator" })
→ "**TokenValidator** (class) — src/auth/token-validator.ts:15
     class TokenValidator { validate(token: string): DecodedToken | null }"

模糊搜索 :

All functions related to payment
→ code_search({ query: "payment", limit: 30 })
→ PaymentProcessor, processPayment, validatePayment, createPaymentIntent, ...

code_context

鉴于任务描述,找到最相关的符号并返回他们的代码,呼叫者,被呼叫者. 没有手动读取文件的编码任务的靴子上下文 .

Args:
  task        string   Task or feature description
  max_symbols number?  Max symbols (default 10)

示例

I need to add rate limiting to the login endpoint
→ code_context({ task: "add rate limiting to login endpoint" })
→ 
### loginHandler (function) — src/routes/auth.ts:67-89
> calledBy: router.post | calls: validateCredentials, createSession
\`\`\`typescript
async function loginHandler(req: Request, res: Response) {
  const { email, password } = req.body;
  ...
}
\`\`\`

特工立即知道哪些文件和符号是相关的.

skills_search

通过查询搜索可用的技能 .

Args:
  query  string   Search term
  cwd    string?  Project directory for stack detection

示例

Find skills for TypeScript testing
→ skills_search({ query: "typescript testing" })
→ "- **ts-jest-setup** ✓ [global, typescript, ACTIVE]
     Configure Jest for TypeScript with proper transforms
   - **vitest-config** [node, typescript]
     Set up Vitest for unit testing"

✓ 表示技能与检测到的堆栈相匹配.

skills_install

从 GitHub 或本地路径安装技能包 .

Args:
  source  string   GitHub URL or local path
  skill   string?  Install only a specific skill
  pack    string?  Override the derived pack name

示例

Install our company skills repo
→ skills_install({ source: "https://github.com/my-org/agent-skills" })
→ "Installed pack "backend" with 12 skill(s)."

skills_list

列出所有已安装的技能包 .

Args: (none)

示例

What packs are installed?
→ skills_list()
→ "**global** (builtin)
     Skills: code-review, refactor-pattern, ...
   **my-org-backend** (installed 2026-06-01)
     Skills: stripe-integration, postgres-pattern, ..."

代码情报索引

索引器使用 植树人 (WASM)将源代码解析为带有符号及其关系的 SQLite 图. 存储位置 :

~/.{BK_APP_NAME}/code-index/{base64url(CWD)}.db

默认 : ~/.deepseek-code/code-index/{hash}.db

所支持的语言: TypeScript, JavaScript, Go, Python, Java, Kotlin, Rust.

通过代理核心连接

import { createCodingEngine } from '@backendkit-labs/agent-coding';

const engine = createCodingEngine({
  providers:       { deepseek: { apiKey: process.env.DEEPSEEK_API_KEY! } },
  defaultProvider: 'deepseek',
  mcpServers: [
    {
      name:    'code-intelligence',
      command: 'bk-mcp',
      args:    [],
      env: {
        BK_APP_NAME: 'my-app',
        BK_CWD:      process.cwd(),
      },
    },
  ],
  subAgentContextMessages: 20,
});

// Index then query
await engine.run('Index this codebase');
await engine.run('Find all functions that directly access the database without using the repository layer');

通过 Claude 桌面连接

添加到您的 Claude 桌面配置 :

马可斯: ~/Library/Application Support/Claude/claude_desktop_config.json
窗口 : %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "bk-code": {
      "command": "bk-mcp",
      "env": {
        "BK_APP_NAME": "claude-code",
        "BK_CWD":      "/path/to/your/project"
      }
    }
  }
}

重新启动克洛德桌面——工具会出现在取取取器中.

多个项目

{
  "mcpServers": {
    "bk-api": {
      "command": "bk-mcp",
      "env": { "BK_CWD": "/projects/api-server", "BK_APP_NAME": "claude-code" }
    },
    "bk-frontend": {
      "command": "bk-mcp",
      "env": { "BK_CWD": "/projects/frontend", "BK_APP_NAME": "claude-code" }
    }
  }
}

HTTP 模式

BK_CWD=/my-project bk-mcp --http --port 3100

连接 :

await mcp.connectStreamableHTTP({
  name:    'code-intelligence',
  baseUrl: 'http://localhost:3100',
});

环境变量