Skip to content

mcp-server.md

Latest commit

 

History

History
354 lines (250 loc) · 7.08 KB

mcp-server.md

File metadata and controls

354 lines (250 loc) · 7.08 KB
Raw

MCP Server Setup Guide

Skill Seekers v3.2.0
Integrate with AI agents via Model Context Protocol

What is MCP?

MCP (Model Context Protocol) lets AI agents like Claude Code control Skill Seekers through natural language:

You: "Scrape the React documentation"
Claude: ▶️ scrape_docs({"url": "https://react.dev/"})
        ✅ Done! Created output/react/

Installation

# Install with MCP support
pip install skill-seekers[mcp]

# Verify
skill-seekers-mcp --version

Transport Modes

stdio Mode (Default)

For Claude Code, VS Code + Cline:

skill-seekers-mcp

Use when:

  • Running in Claude Code
  • Direct integration with terminal-based agents
  • Simple local setup

HTTP Mode

For Cursor, Windsurf, HTTP clients:

# Start HTTP server
skill-seekers-mcp --transport http --port 8765

# Custom host
skill-seekers-mcp --transport http --host 0.0.0.0 --port 8765

Use when:

  • IDE integration (Cursor, Windsurf)
  • Remote access needed
  • Multiple clients

Claude Code Integration

Automatic Setup

# In Claude Code, run:
/claude add-mcp-server skill-seekers

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

{
  "mcpServers": {
    "skill-seekers": {
      "command": "skill-seekers-mcp",
      "env": {
        "ANTHROPIC_API_KEY": "sk-ant-...",
        "GITHUB_TOKEN": "ghp_..."
      }
    }
  }
}

Usage

Once connected, ask Claude:

"List available configs"
"Scrape the Django documentation"
"Package output/react for Gemini"
"Enhance output/my-skill with security-focus workflow"

Cursor IDE Integration

Setup

  1. Start MCP server:
skill-seekers-mcp --transport http --port 8765
  1. In Cursor Settings → MCP:
    • Name: skill-seekers
    • URL: http://localhost:8765

Usage

In Cursor chat:

"Create a skill from the current project"
"Analyze this codebase and generate a cursorrules file"

Windsurf Integration

Setup

  1. Start MCP server:
skill-seekers-mcp --transport http --port 8765
  1. In Windsurf Settings:
    • Add MCP server endpoint: http://localhost:8765

Available Tools

27 tools organized by category:

Core Tools (9)

  • list_configs - List presets
  • generate_config - Create config from URL
  • validate_config - Check config
  • estimate_pages - Page estimation
  • scrape_docs - Scrape documentation
  • package_skill - Package skill
  • upload_skill - Upload to platform
  • enhance_skill - AI enhancement
  • install_skill - Complete workflow

Extended Tools (10)

  • scrape_github - GitHub repo
  • scrape_pdf - PDF extraction
  • scrape_generic - Generic scraper for 10 new source types (see below)
  • scrape_codebase - Local code
  • unified_scrape - Multi-source
  • detect_patterns - Pattern detection
  • extract_test_examples - Test examples
  • build_how_to_guides - How-to guides
  • extract_config_patterns - Config patterns
  • detect_conflicts - Doc/code conflicts

Config Sources (5)

  • add_config_source - Register git source
  • list_config_sources - List sources
  • remove_config_source - Remove source
  • fetch_config - Fetch configs
  • submit_config - Submit configs

Vector DB (4)

  • export_to_weaviate
  • export_to_chroma
  • export_to_faiss
  • export_to_qdrant

scrape_generic Tool

The scrape_generic tool is the generic entry point for 10 new source types added in v3.2.0. It delegates to the appropriate CLI scraper module.

Supported source types: jupyter, html, openapi, asciidoc, pptx, rss, manpage, confluence, notion, chat

Parameters:

Name Type Required Description
source_type string Yes One of the 10 supported source types
name string Yes Skill name for the output
path string No File or directory path (for file-based sources)
url string No URL (for URL-based sources like confluence, notion, rss)

Usage examples:

"Scrape the Jupyter notebook analysis.ipynb"
→ scrape_generic(source_type="jupyter", name="analysis", path="analysis.ipynb")

"Extract content from the API spec"
→ scrape_generic(source_type="openapi", name="my-api", path="api-spec.yaml")

"Process the PowerPoint slides"
→ scrape_generic(source_type="pptx", name="slides", path="presentation.pptx")

"Scrape the Confluence wiki"
→ scrape_generic(source_type="confluence", name="wiki", url="https://wiki.example.com")

See MCP Reference for full details.

Common Workflows

Workflow 1: Documentation Skill

User: "Create a skill from React docs"
Claude: ▶️ scrape_docs({"url": "https://react.dev/"})
        ⏳ Scraping...
        ✅ Created output/react/
        
        ▶️ package_skill({"skill_directory": "output/react/", "target": "claude"})
        ✅ Created output/react-claude.zip
        
        Skill ready! Upload to Claude?

Workflow 2: GitHub Analysis

User: "Analyze the facebook/react repo"
Claude: ▶️ scrape_github({"repo": "facebook/react"})
        ⏳ Analyzing...
        ✅ Created output/react/
        
        ▶️ enhance_skill({"skill_directory": "output/react/", "workflow": "architecture-comprehensive"})
        ✅ Enhanced with architecture analysis

Workflow 3: Multi-Platform Export

User: "Create Django skill for all platforms"
Claude: ▶️ scrape_docs({"config": "django"})
        ✅ Created output/django/
        
        ▶️ package_skill({"skill_directory": "output/django/", "target": "claude"})
        ▶️ package_skill({"skill_directory": "output/django/", "target": "gemini"})
        ▶️ package_skill({"skill_directory": "output/django/", "target": "openai"})
        ✅ Created packages for all platforms

Configuration

Environment Variables

Set in ~/.claude/mcp.json or before starting server:

export ANTHROPIC_API_KEY=sk-ant-...
export GOOGLE_API_KEY=AIza...
export OPENAI_API_KEY=sk-...
export GITHUB_TOKEN=ghp_...

Server Options

# Debug mode
skill-seekers-mcp --verbose

# Custom port
skill-seekers-mcp --port 8080

# Allow all origins (CORS)
skill-seekers-mcp --cors

Security

Local Only (stdio)

# Only accessible by local Claude Code
skill-seekers-mcp

HTTP with Auth

# Use reverse proxy with auth
# nginx, traefik, etc.

API Key Protection

# Don't hardcode keys
# Use environment variables
# Or secret management

Troubleshooting

"Server not found"

# Check if running
curl http://localhost:8765/health

# Restart
skill-seekers-mcp --transport http --port 8765

"Tool not available"

# Check version
skill-seekers-mcp --version

# Update
pip install --upgrade skill-seekers[mcp]

"Connection refused"

# Check port
lsof -i :8765

# Use different port
skill-seekers-mcp --port 8766

See Also