Claude MCP Tools

★ 4

from GrimFandango42

An MCP server ecosystem for integrating with Anthropic's Claude Desktop and Claude Code CLI.

🔥✓ VerifiedAccount requiredAdvanced setup

#automation #analytics #scraping #browser #database #documents #devops #git

⚠️ ARCHIVED — Claude-MCP-tools

This repository has been archived. All functionality has been superseded by OpenClaw’s native skill ecosystem, Desktop Commander, and community MCP connectors.

Routing concept (AgenticSeek): The smart provider-routing idea lives on as a design reference for future OpenClaw cost-aware routing features.

Decision date: 2026-02-14 Reason: 14 MCP servers evaluated; none provide unique value beyond existing tooling.

Claude MCP Tools

Professional Model Context Protocol (MCP) server ecosystem for Anthropic's Claude Desktop and Claude Code CLI integration.

🎯 Current Status: Fully Tested Production-Ready Multi-Provider AI System ✅

Latest Update: June 2, 2025
Testing Status: ⭐ COMPREHENSIVE TESTING COMPLETE - 94/100 overall score
Server Count: 21 operational MCP servers (consolidated from 25)
Tool Validation: All 7 core MCP tool categories verified working perfectly
Performance: Sub-second response times across all tools
Integration: 3/3 complex workflow scenarios tested and functional
System Status: Claude Code Integration tested with 100% success rate
Achievement: Comprehensive memory system and enhanced context retention

Overview

This repository provides a production-ready ecosystem of 21 MCP servers enabling advanced AI-assisted development workflows. The servers facilitate Claude's interaction with external systems, APIs, and development tools through standardized Model Context Protocol implementations.

Key Capabilities

Hybrid AI Development: Seamless integration between Claude Desktop and Claude Code CLI
Multi-Provider AI Routing: Smart routing between local and cloud AI providers with cost optimization
Advanced Memory Management: Persistent context storage with rich entity-relation graphs
Desktop Automation: Comprehensive GUI automation and testing capabilities
Development Workflow Enhancement: Code analysis, formatting, security scanning, and project management

Architecture

Server Ecosystem (21 Servers)

Production Custom Servers (12)

Claude Code Integration MCP ⭐ TESTED & VERIFIED - Task delegation and project analysis
AgenticSeek MCP - Multi-provider AI routing (Local DeepSeek, OpenAI, Google Gemini)
Vibetest MCP - Multi-agent browser QA testing swarm
Code Formatter MCP - Black/Prettier wrapper for code formatting
Security Scanner MCP - Vulnerability scanning with pip-audit and npm audit
Windows Computer Use MCP - Desktop automation with screen capture
Containerized Computer Use MCP - Docker-isolated GUI automation
API Gateway MCP - Unified routing for OpenAI and Anthropic APIs
Financial Datasets MCP - Financial data integration and analysis
N8n Workflow Generator MCP - Workflow automation platform
Docker Orchestration MCP - Container lifecycle management
Knowledge Memory MCP - Vector-based persistent storage

Integrated Third-Party Servers (9)

GitHub Integration MCP - Repository management and automation
Firecrawl Custom MCP - Web scraping and content extraction
ScreenPilot MCP - Advanced UI element detection and automation
SQLite MCP - Local database operations
Memory MCP (Official) - Standard MCP memory implementation
Filesystem MCP - Secure file operations with path sandboxing
Sequential Thinking MCP - Structured problem-solving framework
Playwright MCP - Browser automation and testing
Fantasy Premier League MCP - Sports analytics and data

Platform Compatibility

Server Category	Claude Desktop	Claude Code	Notes
AI Routing	✅	✅	Multi-provider smart routing
Memory Management	✅	✅	Shared persistent context
Code Intelligence	✅	✅	Analysis, formatting, security
Desktop Automation	✅	❌	Requires GUI environment
API Services	✅	✅	REST/GraphQL integration
Container Management	✅	✅	Docker orchestration

Core Features

Memory-First Architecture

All servers integrate with the Memory MCP system for persistent context:

Entity Storage: Projects, servers, processes, and knowledge
Relation Mapping: Connected information discovery
Cross-Session Continuity: Context preservation across conversations
Searchable Knowledge: Single-keyword search with comprehensive results

Smart AI Routing

AgenticSeek MCP provides intelligent AI provider selection:

Local Processing: Free, private DeepSeek AI (privacy priority)
Cloud APIs: OpenAI GPT-3.5/4 (speed priority) and Google Gemini (cost priority)
Smart Routing: Automatic provider selection based on task characteristics
Cost Optimization: Transparent cost estimation and optimization

Development Workflow Enhancement

Comprehensive tools for AI-assisted development:

Code Analysis: AST parsing, complexity metrics, symbol resolution
Quality Assurance: Automated formatting, linting, and security scanning
Task Delegation: Claude Code Integration for complex coding tasks
Project Intelligence: Automated codebase analysis and recommendations

Server Implementation

Basic Server Pattern

from fastmcp import FastMCP

mcp = FastMCP("server-name")

@mcp.tool()
async def process_data(input: str, options: dict = None) -> dict:
    """Process data with specified options."""
    # Server implementation
    return {"status": "success", "result": processed_data}

if __name__ == "__main__":
    mcp.run(transport="stdio")

Configuration Schema

{
  "mcpServers": {
    "server-name": {
      "command": "python",
      "args": ["path/to/server.py"],
      "cwd": "working/directory",
      "env": {
        "API_KEY": "your-api-key",
        "LOG_LEVEL": "INFO"
      },
      "keepAlive": true,
      "stderrToConsole": true
    }
  }
}

Development Guidelines

Server Development Standards

FastMCP Framework: Use for all new MCP servers
Async Patterns: Proper async/await implementation
Error Handling: Comprehensive error management and logging
Memory Integration: Document discoveries in Memory MCP
Testing: Independent testing before Claude Desktop integration

Code Quality Standards

Logging: Structured logging to stderr only (stdout reserved for JSON-RPC)
Type Hints: Full type annotation for better tooling
Documentation: Rich docstrings for tools and functions
Security: Input validation and secure API key management

Recent Achievements

June 2, 2025 - Comprehensive MCP Testing & Validation Complete ✅

Full System Testing: 94/100 overall score with all 7 core tool categories working perfectly
Live Tool Validation: Memory (27 entities), AgenticSeek (4 AI providers), Firecrawl, GitHub, SQLite operational
Integration Workflows: 3/3 complex scenarios tested and verified functional
Performance Excellence: Sub-second response times across all MCP tools
Testing Framework: Automated testing suite created for continuous validation
Comprehensive Memory Structure: 27 entities, 31 relations for project knowledge
Claude Code Integration Testing: 100% success rate (2/2 tasks completed)
Enhanced Documentation: Professional system instructions and workflow templates
Server Consolidation: Optimized from 25 to 21 servers with improved efficiency

May 31, 2025 - AI Routing System Complete ✅

AgenticSeek MCP Server: AsyncIO bug resolution with 100% operational success
Multi-Provider Integration: Local DeepSeek, OpenAI GPT-3.5/4, Google Gemini
Smart Cost Optimization: Intelligent provider selection based on task characteristics

Project Structure

Claude-MCP-tools/
├── servers/                    # MCP server implementations
│   ├── agenticseek-mcp/       # Multi-provider AI routing
│   ├── claude-code-integration-mcp/  # Claude Code CLI bridge
│   ├── windows-computer-use/   # Desktop automation
│   └── ...                    # Additional servers
├── docs/                      # Comprehensive documentation
├── scripts/                   # Deployment and testing scripts
├── archive/                   # Historical and deprecated files
├── CLAUDE.md                  # Development guidelines
├── PROJECT_STATUS_UPDATED.md  # Current project status
└── CONFIG_UPDATE_COMPLETE.md  # Configuration management guide

Support & Resources

Documentation: Comprehensive guides in /docs/ directory
Development Guidelines: See CLAUDE.md for standards and patterns
Project Status: Current status in PROJECT_STATUS_UPDATED.md
Memory Workflow: MEMORY_WORKFLOW_GUIDE.md for context management
Session Templates: SESSION_CONTEXT_TEMPLATE.md for structured work

Copy & paste — that's it

# Install FastMCP framework
pip install fastmcp python-json-logger

# Install development tools (optional)
pip install black prettier pip-audit safety

Before it works, you'll need: API_KEYLOG_LEVEL

Quick Start

Prerequisites

Python 3.11+ or Node.js 18+
Claude Desktop or Claude Code CLI
FastMCP framework for custom server development

Installation

Clone the repository:

git clone https://github.com/GrimFandango42/Claude-MCP-tools.git
cd Claude-MCP-tools

Install core dependencies:

# Install FastMCP framework
pip install fastmcp python-json-logger

# Install development tools (optional)
pip install black prettier pip-audit safety

Configure Claude Desktop:

# Copy configuration template
cp claude_desktop_config_template.json $CLAUDE_CONFIG_PATH

# Edit configuration with your API keys and paths
# Location: %APPDATA%\Claude\claude_desktop_config.json (Windows)

Restart Claude Desktop to load servers

Verification

Test your setup by asking Claude:

"Search memory for server status and show me what MCP servers are available"

Memory System Usage

Session Workflow

Start sessions by searching memory for relevant context
Document discoveries in memory during work
Create relations between connected components
End sessions by updating memory with outcomes

Effective Search Patterns

# Search for project context
mcp__memory__search_nodes("server")
mcp__memory__search_nodes("consolidation") 
mcp__memory__search_nodes("FastMCP")

# Search by entity type
mcp__memory__search_nodes("configuration")
mcp__memory__search_nodes("documentation")

Troubleshooting

Common Issues

Issue	Solution
Server not connecting	Check logs at `%APPDATA%\Claude\logs\mcp-server-{name}.log`
FastMCP import errors	Install to system site-packages: `pip install --target system fastmcp`
Memory search inconsistency	Use single keywords instead of multi-word phrases
Configuration not loading	Restart Claude Desktop after config changes
AsyncIO event loop conflicts	Use proper async/await patterns, avoid `asyncio.run()`

Debug Configuration

{
  "stderrToConsole": true,
  "env": {
    "MCP_LOG_LEVEL": "DEBUG",
    "CLAUDE_CODE_MOCK": "true"
  }
}

⚠️ ARCHIVED — Claude-MCP-tools

Claude MCP Tools

🎯 Current Status: Fully Tested Production-Ready Multi-Provider AI System ✅

Overview

Key Capabilities

Architecture

Server Ecosystem (21 Servers)

Production Custom Servers (12)

Integrated Third-Party Servers (9)

Platform Compatibility

Core Features

Memory-First Architecture

Smart AI Routing

Development Workflow Enhancement

Server Implementation

Basic Server Pattern

Configuration Schema

Development Guidelines

Server Development Standards

Code Quality Standards

Recent Achievements

June 2, 2025 - Comprehensive MCP Testing & Validation Complete ✅

May 31, 2025 - AI Routing System Complete ✅

Project Structure

Support & Resources

Quick Start

Prerequisites

Installation

Verification

Memory System Usage

Session Workflow

Effective Search Patterns

Troubleshooting

Common Issues

Debug Configuration

License