Better Finder MCP

Better Finder MCP

Enables intelligent file search with Git-like staging and indexing, offering semantic and hybrid search for documents, and integrates with Claude Desktop via MCP.

Category
访问服务器

README

🔍 Better Finder MCP

Git-like workflow for intelligent file search and Claude MCP integration

Better Finder transforms file discovery with semantic search, hybrid matching, and a familiar Git-style workflow. Index your documents, search with natural language, and integrate seamlessly with Claude Desktop.

🚀 Quick Start

Installation

# Clone and install
git clone https://github.com/GitHamza0206/better-finder-mcp.git
cd better-finder-mcp
uv venv
source .venv/bin/activate
pip install -e .

# Install it globally with uv
uv tool install -e .

Basic Workflow

# 1. Stage files for indexing (like git add)
better-finder add ~/Documents

# 2. Check what's staged
better-finder status

# 3. Index staged files
better-finder index

# 4. Search your files
better-finder search "quarterly financial reports"

📋 Commands

File Staging

  • better-finder add <path> - Stage files or Documents for indexing
  • better-finder rm <path> - Remove files from staging
  • better-finder status - Show staged files
  • better-finder index - Index staged files

Search & Management

  • better-finder search <query> - Search indexed files
  • better-finder stats - Show index statistics
  • better-finder clear-index - Clear all indexed data
  • better-finder server - Start MCP server for Claude

Utilities

  • better-finder show <file> - Display file content
  • better-finder remove-file <file> - Remove file from index

🎯 Key Features

Hybrid Search

Combines semantic search, keyword matching, and fuzzy filename search for comprehensive results.

Git-like Workflow

Familiar staging process gives you precise control over what gets indexed.

.betterfinderignore Support

Use ignore patterns to exclude sensitive files:

# .betterfinderignore
secrets/
*.key
temp-*.pdf
node_modules/

Supported File Types

  • Documents: PDF, DOC, DOCX, TXT, MD, RTF, ODT
  • Spreadsheets: XLSX, XLS, CSV, ODS
  • Presentations: PPTX, PPT
  • Data: JSON, XML

Claude MCP Integration

Start the MCP server to use Better Finder directly within Claude Desktop:

better-finder server

Add to your Claude Desktop config (~/.cursor/mcp.json):

{
  "mcpServers": {
    "better-finder": {
      "command": "better-finder",
      "args": ["server"]
    }
  }
}

🔧 Configuration

Better Finder uses smart defaults:

  • Default scan paths: ~/Documents, ~/Desktop, ~/Downloads
  • Ignored directories: .git, node_modules, .venv, cache folders
  • Chunk size: 2000 characters for better context
  • Search threshold: 0.4 for comprehensive results

💡 Examples

Document Discovery

# Find presentations about sales
better-finder search "sales presentation Q4"

# Look for specific file types
better-finder search "budget" --type excel

# Get more results
better-finder search "meeting notes" --max 20

Staging Workflow

# Stage entire project documentation
better-finder add ./docs

# Remove sensitive files
better-finder rm ./docs/secrets/

# Check what will be indexed
better-finder status

# Index everything staged
better-finder index

MCP Integration

Once the server is running, ask Claude:

  • "Search my documents for budget reports"
  • "Find presentations about project timelines"
  • "Show me files related to client proposals"

🏗️ Architecture

  • File Processors: Extract content from different formats
  • FAISS Vector Store: Fast similarity search with sentence transformers
  • SQLite Metadata: Efficient file metadata and chunk storage
  • Staging System: Git-like file management with JSON persistence
  • MCP Server: Model Context Protocol integration for Claude

📊 Performance

  • Indexing: ~100-500 files per minute
  • Search: Sub-second results
  • Memory: ~200-500 MB for typical collections
  • Storage: ~1-5 MB per 1000 documents

📝 License

MIT License - see LICENSE file for details.


Made for developers who want intelligent file discovery with familiar Git-like controls.

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