SCAPO MCP Server

A Model Context Protocol (MCP) server that makes your locally-extracted SCAPO knowledge base queryable.

⚠️ This is a reader, not a scraper! You must first use SCAPO to extract tips into your models/ folder.

Documentation

For comprehensive usage instructions, examples, and technical details, please see the Usage Guide.

Prerequisites

1. Clone and set up SCAPO first:

   git clone https://github.com/czero-cc/scapo.git
   cd scapo
   # Follow SCAPO setup to run scrapers and populate models/

2. Required:

   • Node.js 18+
   • npm or npx
   • Populated models/ directory (from running SCAPO scrapers)

How It Works

IMPORTANT: This MCP server ONLY reads from your local models/ folder. It does NOT scrape data itself!

1. First, use SCAPO to scrape and extract tips into models/
2. Then, this MCP server makes those tips queryable in your AI client

Quick Start

# Step 1: Set up SCAPO and extract tips
git clone https://github.com/czero-cc/scapo.git
cd scapo
# Follow SCAPO README to configure and run scrapers
scapo scrape targeted --service "GitHub Copilot" --limit 20

# Step 2: Configure MCP to read your extracted tips
# Add to your MCP client config with YOUR path to scapo/models/

Installation

npx @arahangua/scapo-mcp-server

Configuration for MCP Clients

Add this to your MCP client's configuration:

{
  "mcpServers": {
    "scapo": {
      "command": "npx",
      "args": ["@arahangua/scapo-mcp-server"],
      "env": {
        "SCAPO_MODELS_PATH": "/absolute/path/to/your/scapo/models"  // From your cloned SCAPO repo!
      }
    }
  }
}

Note: Set SCAPO_MODELS_PATH to the absolute path of your SCAPO models directory.

For Claude Desktop specifically:

• Windows: Edit %APPDATA%\Claude\claude_desktop_config.json
• macOS: Edit ~/Library/Application Support/Claude/claude_desktop_config.json

Available Tools

1. get_best_practices

Get AI/ML best practices for a specific model.

Arguments:
- model_name: Model name (e.g., "Qwen3-Coder-Flash", "Llama-3.2-1B")
- practice_type: Type of practices ("all", "prompting", "parameters", "pitfalls")

Example in Claude:

"Can you get me the best practices for Qwen3-Coder-Flash?"

2. search_models

Search for models by keyword.

Arguments:
- query: Search query
- limit: Maximum results (default: 10)

Example in Claude:

"Search for models that are good for coding"

3. list_models

List all available models by category.

Arguments:
- category: Model category ("text", "image", "video", "audio", "multimodal", "code", "all")

Example in Claude:

"List all available text models"

4. get_recommended_models

Get recommended models for a specific use case.

Arguments:
- use_case: Use case (e.g., "code_generation", "creative_writing", "image_generation")

Example in Claude:

"What models do you recommend for code generation?"

Environment Variables

• SCAPO_MODELS_PATH: Path to local models directory (defaults to ../models relative to MCP server)
• SCAPO_API_URL: Optional API endpoint (not needed for basic usage)

Features

• Intelligent Fuzzy Matching: Handles typos, partial names, and variations automatically
  • Typo tolerance: heygen → "HeyGen", gemeni → "Gemini"
  • Partial matching: qwen → finds all Qwen variants
  • Case insensitive: LLAMA-3 → "llama-3"
• Fully Standalone: Works without any API server running
• Direct File Access: Reads from local model files
• Smart Search: Advanced search with similarity scoring
• Smart Recommendations: Suggests models based on use case
• Easy Integration: Works with any MCP-compatible client
• Helpful Suggestions: Provides alternatives when exact matches aren't found

Use Cases

The MCP server recognizes these use cases for recommendations:

• code_generation: Programming and code completion
• creative_writing: Stories, articles, creative content
• image_generation: Text-to-image generation
• chat_conversation: Conversational AI

Directory Structure

The server expects this structure in your models directory:

models/
├── text/
│   ├── Qwen3-Coder-Flash/
│   │   ├── prompting.md
│   │   ├── parameters.json
│   │   ├── pitfalls.md
│   │   └── metadata.json
│   └── Llama-3.2-1B/
│       └── ...
├── image/
│   └── stable-diffusion/
│       └── ...
└── ...

Contributing

To contribute improvements:

1. Fork the SCAPO repository
2. Make your changes in the mcp/ directory
3. Submit a pull request

License

Same as the parent SCAPO repository.