oldnordic/LTMC

LTMC - Long-Term Memory and Context MCP Server

⚠️ PROTOTYPE STATUS - NOT PRODUCTION READY

🔴 This is experimental prototype software - NOT for production use

• Current Status: Prototype/Experimental implementation
• Stability: Breaking changes expected frequently
• Testing: Limited testing in controlled environments only
• Support: No production support or guarantees
• Recommendation: Use only for research and experimentation

Version: 4.0 (Prototype)
Status: 🚧 Experimental Prototype - Architectural Consolidation Phase
Tools: 11 Consolidated MCP Tools (Prototype Implementation) Transport: stdio MCP protocol (Experimental)

🎯 Overview

LTMC is an experimental Model Context Protocol (MCP) server prototype that consolidates memory, context management, and agent coordination capabilities. This prototype demonstrates multi-database architecture integration but is NOT ready for production deployment.

⚠️ This software is provided for experimental and research purposes only.

🏆 Major Achievement

✅ ARCHITECTURAL CONSOLIDATION SUCCESS

• Before: 126+ @mcp.tool decorators scattered across 15+ files
• After: 11 consolidated, comprehensive tools in a single maintainable file
• Improvement: 91.3% complexity reduction while maintaining full functionality
• Quality: Zero shortcuts, mocks, or placeholders - all real implementations

✨ Key Features

• 🧠 11 Consolidated MCP Tools - Complete functionality with optimal maintainability
• 💾 4-Database Integration - SQLite + FAISS + Redis + Neo4j working seamlessly
• 🤖 Enterprise Agent Coordination - Real-time multi-agent workflow orchestration
• 🔍 Advanced Search - Semantic, graph, and hybrid search capabilities
• 📚 Knowledge Graphs - Automatic relationship building with Neo4j
• 🎯 Intelligent Task Management - ML-enhanced complexity analysis
• ⚡ Performance Excellence - All operations <2s SLA, most <500ms
• 🔧 Quality Standards - >94% test coverage, real database operations

🛠️ Technology Stack

Core Technologies:

• Python 3.9+ with asyncio patterns and type hints
• MCP stdio Protocol - Optimized for Claude Code integration
• Multi-Database Architecture:
  • SQLite - Primary data storage with WAL journaling
  • Neo4j - Knowledge graph relationships (<25ms queries)
  • Redis - Real-time caching and coordination (<1ms operations)
  • FAISS - Vector similarity search (<25ms searches)

Quality & Performance:

• Real Implementations Only - No mocks, stubs, or placeholders
• Performance Monitoring - SLA compliance tracking
• Comprehensive Testing - Integration tests with real databases
• Documentation-First - Complete user and technical guides

🚀 Quick Start

1. Installation

git clone https://github.com/oldnordic/ltmc.git
cd ltmc
pip install -r config/requirements.txt

2. Configuration

# Copy example configuration
cp config/ltmc_config.env.example config/ltmc_config.env
# Edit configuration as needed

3. Claude Code Integration

Add to your Claude Code MCP configuration:

{
  "ltmc": {
    "command": "python",
    "args": ["-m", "ltms"],
    "cwd": "/path/to/ltmc"
  }
}

4. Verification

# Test system health
python -c "from ltms.tools.memory.memory_actions import memory_action; print(memory_action(action='status'))"

📚 Documentation

Quick Start

• 📖 - Complete setup instructions
• ⚙️ - Environment setup
• 🎯 - Practical usage examples

Tool Reference

• 🛠️ - Detailed tool documentation
• 🔧 - Multi-agent workflows

Technical Documentation

• 🏗️ - Deep technical dive
• 🎼 - Agent coordination details
• 📊 - System health and metrics
• 📋 - Production deployment

Project Documentation

• 🎯 - Consolidation achievement summary
• 📂 - Complete documentation index

🔧 The 11 Consolidated Tools

🎯 Use Cases

• 🧠 Persistent Memory - Never lose context across conversations
• 🤖 Agent Coordination - Enterprise-grade multi-agent workflows
• 📊 Knowledge Management - Build and query knowledge graphs
• 🔍 Pattern Recognition - Learn from code patterns and experiences
• 📝 Documentation Sync - Keep docs synchronized with code changes
• ⚡ Performance Optimization - Intelligent caching and monitoring

🌟 Why LTMC?

Architectural Excellence

• Successful consolidation from 126+ tools to 11 comprehensive tools
• Quality-over-speed development with real implementations only
• Multi-database integration with transaction-like consistency
• Enterprise-grade agent coordination and workflow management

Prototype Characteristics

• Experimental implementation - Variable performance, no guaranteed SLAs
• Database integration testing - Real database operations in prototype environment
• Limited testing coverage - Controlled environment testing only
• Development monitoring - Basic health checks and metrics (prototype level)

📊 System Status

⚠️ PROTOTYPE STATUS - Experimental Implementation

• Architecture Quality: Prototype - Experimental consolidation
• Performance: Variable - No production SLAs guaranteed
• Code Quality: Experimental - Subject to frequent changes
• Documentation: In Progress - Prototype documentation
• Testing: Limited - Controlled environment testing only

Prototype Metrics (No Production Guarantees):

• Tool Response Time: Variable performance (prototype)
• Database Operations: Experimental implementation
• System Stability: Prototype - Breaking changes expected
• Memory Usage: Not optimized for production

🤝 Contributing

LTMC follows quality-over-speed principles. Please review:

• for system understanding
• for development priorities
• Quality standards: No mocks/stubs, real implementations only

🔗 Links

• GitHub: https://github.com/oldnordic/ltmc
• Documentation:
• MCP Protocol: Model Context Protocol
• System Status:

⚠️ LTMC is an experimental prototype - demonstrating architectural consolidation concepts and multi-database integration patterns. This software is NOT ready for production deployment and should only be used for research and experimental purposes.

Use at your own risk - No warranties or support provided.