99hansling/codeweaver-mcp-server

CodeWeaver MCP Server

Transform your entire codebase into a single, AI-readable Markdown document through the Model Context Protocol.

English |

CodeWeaver MCP Server brings powerful code packaging capabilities to AI assistants like Claude Code. It converts complex project structures into clean, navigable Markdown files - perfect for code reviews, documentation, AI analysis, and sharing with teams.

Why CodeWeaver MCP?

🚀 Seamless AI Integration

• Native MCP Protocol: Works out-of-the-box with Claude Code and other MCP-compatible AI assistants
• Zero Context Switching: Access codebase packaging directly from your AI workflow
• Instant Results: Get Markdown output immediately or save to file

📦 Complete Codebase Visibility

• Full Structure Capture: Beautiful tree-view of your entire project hierarchy
• All Code Included: Every file's content embedded with proper syntax highlighting
• Smart Filtering: Respects .gitignore and supports custom ignore patterns

💡 Perfect for Multiple Use Cases

For AI Analysis

• Feed entire codebases to AI models (ChatGPT, Claude, etc.)
• Enable comprehensive code understanding without manual file sharing
• Perfect for code reviews and refactoring discussions

For Documentation

• Generate instant project snapshots
• Create shareable codebase documentation
• Archive project states at key milestones

For Team Collaboration

• Share complete project context with teammates
• Onboard new developers faster
• Facilitate remote code discussions

For Code Reviews

• Package pull requests with full context
• Review entire features in one document
• Track code changes across multiple files

Features

🎯 Pure TypeScript Implementation

• No External Dependencies: Built entirely in TypeScript - no Go, Python, or other runtimes required
• Fast & Lightweight: Efficient file scanning and processing
• Cross-Platform: Works on macOS, Linux, and Windows

🛠️ Powerful Customization

• Flexible Filtering: Include/exclude files with glob patterns
• Gitignore Support: Automatically respects your .gitignore rules
• Custom Patterns: Add project-specific ignore patterns
• Multiple Output Modes: Return content directly or save to file

📝 Smart Markdown Output

• Syntax Highlighting: Automatic language detection for 20+ programming languages
• Clean Formatting: Professional tree structure using box-drawing characters
• Organized Layout: Logical file ordering (directories first, alphabetical)
• Relative Paths: Clear file location references

Installation

Prerequisites

• Node.js 18 or later
• npm or yarn

Quick Install

# Clone and build
git clone https://github.com/yourusername/codeweaver-mcp-server.git
cd codeweaver-mcp-server
npm install
npm run build

Add to Claude Code

claude mcp add codeweaver node /absolute/path/to/codeweaver-mcp-server/dist/index.js

Or manually add to ~/.claude/config.json:

{
  "mcpServers": {
    "codeweaver": {
      "command": "node",
      "args": ["/absolute/path/to/codeweaver-mcp-server/dist/index.js"]
    }
  }
}

Usage

Basic Usage

Pack current directory:

codeweaver_pack_codebase({
  path: "."
})

Save to File

codeweaver_pack_codebase({
  path: "./src",
  output: "codebase.md"
})

Include Gitignored Files

codeweaver_pack_codebase({
  path: ".",
  include_gitignored: true
})

Custom Ignore Patterns

codeweaver_pack_codebase({
  path: ".",
  ignore_patterns: ["*.log", "tmp/*", "*.tmp"]
})

Output Format

CodeWeaver generates clean, structured Markdown:

# project-name

## Directory Structure

├── src │ ├── index.ts │ └── utils.ts └── package.json

## Files

### src/index.ts

```typescript
// Full file content with syntax highlighting
import { helper } from './utils';
...

src/utils.ts

// Full file content
export function helper() { ... }

## Supported Languages

Automatic syntax highlighting for:
- TypeScript/JavaScript (`.ts`, `.tsx`, `.js`, `.jsx`)
- Python (`.py`)
- Go (`.go`)
- Rust (`.rs`)
- Java (`.java`)
- C/C++ (`.c`, `.cpp`, `.h`, `.hpp`)
- C# (`.cs`)
- Ruby (`.rb`)
- PHP (`.php`)
- Swift (`.swift`)
- Kotlin (`.kt`)
- SQL (`.sql`)
- Shell (`.sh`)
- YAML (`.yaml`, `.yml`)
- JSON (`.json`)
- XML (`.xml`)
- HTML (`.html`)
- CSS (`.css`)
- Markdown (`.md`)

## Default Ignore Patterns

CodeWeaver automatically skips:
- `node_modules/`
- `.git/`
- `dist/` and `build/`
- `.next/` and `.nuxt/`
- `vendor/`
- Python virtual environments (`venv/`, `.venv/`, `__pycache__/`)
- System files (`.DS_Store`, `Thumbs.db`)

## Real-World Examples

### Example 1: Share Code with ChatGPT
```typescript
// Pack your project
codeweaver_pack_codebase({
  path: "./my-app",
  output: "my-app-context.md"
})

// Copy the Markdown file content and paste into ChatGPT
// Now ChatGPT can see your entire codebase!

Example 2: Code Review Documentation

// Pack only source code
codeweaver_pack_codebase({
  path: "./src",
  ignore_patterns: ["*.test.ts", "*.spec.ts"],
  output: "review-package.md"
})

Example 3: Project Snapshot

// Capture complete project state
codeweaver_pack_codebase({
  path: ".",
  output: `snapshot-${new Date().toISOString()}.md`
})

API Reference

codeweaver_pack_codebase

Parameters:

Returns:

• If output specified: Success message with file path and size
• If output omitted: Complete Markdown content

Error Handling:

• "Error: Path does not exist" - Invalid directory path
• "Error: Path is not a directory" - Path points to a file
• Binary files automatically skipped with notice

Development

Build from Source

npm install
npm run build

Development Mode

npm run dev

Clean Build

npm run clean
npm run build

Why Choose CodeWeaver MCP?

✅ For AI-First Workflows: Built specifically for AI tools, unlike generic file packers ✅ Zero Configuration: Works immediately with sensible defaults ✅ Professional Output: Clean, readable Markdown that humans and AI love ✅ Modern Stack: Pure TypeScript, no legacy dependencies ✅ Active Development: Regularly updated with new features ✅ MIT Licensed: Free for personal and commercial use

Comparison with Alternatives

Contributing

Contributions welcome! Please feel free to submit issues or pull requests.

License

MIT License - see file for details

Acknowledgments

Inspired by the original CodeWeaver project and adapted for the Model Context Protocol ecosystem.

Made with ❤️ for the AI development community