MCP Codebase Index

MCP Codebase Index

Enables semantic search across your codebase using Google's Gemini embeddings and Qdrant Cloud vector storage. Supports 15+ programming languages with smart code chunking and real-time file change monitoring.

Category
访问服务器

README

MCP Codebase Index Server

AI-powered semantic search for your codebase in Claude Desktop

A Model Context Protocol (MCP) server that enables Claude to search and understand your codebase using Google's Gemini embeddings and Qdrant Cloud vector storage.

✨ Features

  • 🔍 Semantic Search: Find code by meaning, not just keywords
  • 🎯 Smart Chunking: Automatically splits code into logical functions/classes
  • 🔄 Real-time Watch: Monitors file changes and updates index automatically
  • 🌐 Multi-language: Supports 15+ programming languages
  • ☁️ Cloud Storage: Uses Qdrant Cloud for persistent vector storage
  • 📦 Simple Setup: Just 4 environment variables to get started

🚀 Quick Start

Prerequisites

  1. Gemini API Key: Get free at Google AI Studio
  2. Qdrant Cloud Account: Sign up free at cloud.qdrant.io

Installation

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "codebase": {
      "command": "npx",
      "args": ["-y", "@ngotaico/mcp-codebase-index"],
      "env": {
        "REPO_PATH": "/absolute/path/to/your/project",
        "GEMINI_API_KEY": "AIzaSyC...",
        "QDRANT_URL": "https://your-cluster.gcp.cloud.qdrant.io:6333",
        "QDRANT_API_KEY": "eyJhbGci..."
      }
    }
  }
}

All 4 variables are required:

Variable Where to Get Example
REPO_PATH Absolute path to your project /Users/you/Projects/myapp
GEMINI_API_KEY Google AI Studio AIzaSyC...
QDRANT_URL Qdrant Cloud cluster URL https://xxx.gcp.cloud.qdrant.io:6333
QDRANT_API_KEY Qdrant Cloud API key eyJhbGci...

Restart Claude Desktop

The server will automatically:

  1. Connect to your Qdrant Cloud cluster
  2. Create a collection (if needed)
  3. Index your entire codebase
  4. Watch for file changes

📖 Usage

Ask Claude to search your codebase:

"Find the authentication logic"
"Show me how database connections are handled"  
"Where is error logging implemented?"
"Find all API endpoint definitions"

🎛️ Configuration

Required Variables

{
  "env": {
    "REPO_PATH": "/Users/you/Projects/myapp",
    "GEMINI_API_KEY": "AIzaSyC...",
    "QDRANT_URL": "https://xxx.gcp.cloud.qdrant.io:6333",
    "QDRANT_API_KEY": "eyJhbGci..."
  }
}

Optional Variables

{
  "env": {
    "QDRANT_COLLECTION": "my_project",
    "WATCH_MODE": "true",
    "BATCH_SIZE": "50"
  }
}
Variable Default Description
QDRANT_COLLECTION codebase Collection name in Qdrant
WATCH_MODE true Auto-update on file changes
BATCH_SIZE 50 Embedding batch size

🔧 Setup Guides

🌍 Supported Languages

Python • TypeScript • JavaScript • Dart • Go • Rust • Java • Kotlin • Swift • Ruby • PHP • C • C++ • C# • Shell • SQL • HTML • CSS

📝 How It Works

┌─────────────┐
│  Your Code  │
└──────┬──────┘
       │
       ▼
┌─────────────────┐
│  File Watcher   │  Monitors changes
└──────┬──────────┘
       │
       ▼
┌─────────────────┐
│  Code Parser    │  Splits into chunks
└──────┬──────────┘
       │
       ▼
┌─────────────────┐
│  Gemini API     │  Creates embeddings
└──────┬──────────┘
       │
       ▼
┌─────────────────┐
│  Qdrant Cloud   │  Stores vectors
└──────┬──────────┘
       │
       ▼
┌─────────────────┐
│  Claude Search  │  Semantic queries
└─────────────────┘

🐛 Troubleshooting

Server not appearing in Claude?

Check Claude logs:

tail -f ~/Library/Logs/Claude/mcp*.log

Common issues:

  • REPO_PATH must be absolute path
  • ❌ All 4 env variables must be set
  • ❌ Qdrant URL must include :6333 port
  • ❌ Gemini API key must be valid

Can't connect to Qdrant?

Test connection:

curl -H "api-key: YOUR_KEY" \
  https://YOUR_CLUSTER.gcp.cloud.qdrant.io:6333/collections

Should return JSON with collections list.

Indexing too slow?

  • Large repos (1000+ files) take 5-10 minutes initially
  • Reduce BATCH_SIZE if hitting rate limits
  • Check Gemini API quota: aistudio.google.com

📊 Performance

  • Embedding speed: ~100 chunks/minute (Gemini API)
  • Search latency: <100ms (Qdrant Cloud)
  • Storage: ~1KB per code chunk
  • Recommended: <10K chunks per collection

📄 License

MIT © NgoTaiCo

🤝 Contributing

Issues and PRs welcome at github.com/NgoTaiCo/mcp-codebase-index

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选