MCP-Context-Provider

MCP Context Provider

A static MCP (Model Context Protocol) server that provides AI models with persistent tool context, preventing context loss between chat sessions. This server automatically loads and injects tool-specific rules, syntax preferences, and best practices at Claude Desktop startup.

Overview

The Context Provider acts as a persistent neural core for your AI interactions, eliminating the need to re-establish context in each new chat session by:

• 🔄 Persistent Context: Like the stable orb in the visualization, rules and preferences survive across Claude Desktop restarts
• ⚡ Automatic Injection: Context flows seamlessly into every conversation, just as the data streams connect to the central core
• 🎯 Tool-Specific: Each tool gets its own context rules and syntax preferences, creating specialized knowledge pathways
• 🔧 Auto-Corrections: Automatic syntax transformations (e.g., Markdown → DokuWiki) ensure consistency across all interactions
• 📈 Scalable: Easy to add new tools and context rules, expanding the knowledge network
• 🏢 Enterprise-Ready: Version-controlled context management provides organizational stability

The Neural Network Metaphor

Just like the image depicts, your MCP Context Provider functions as:

• Central Orb: The stable, persistent context core that maintains consistency
• Neural Pathways: Tool-specific context rules that create specialized knowledge channels
• Data Streams: Individual chat sessions that flow through and benefit from the persistent context
• Network Stability: Prevents the ephemeral nature of conversations from losing important contextual information

Quick Start

Option 1: Automated Installation (Recommended)

The easiest way to install MCP Context Provider is using the provided installation scripts:

Unix/Linux/macOS:

# Download the DXT package
wget https://github.com/doobidoo/MCP-Context-Provider/raw/main/mcp-context-provider-1.1.0.dxt

# Run the installation script
curl -sSL https://raw.githubusercontent.com/doobidoo/MCP-Context-Provider/main/install.sh | bash

Windows:

# Download and run the Windows installer
Invoke-WebRequest -Uri "https://raw.githubusercontent.com/doobidoo/MCP-Context-Provider/main/install.bat" -OutFile "install.bat"
.\install.bat

The installation script automatically:

• Unpacks the DXT extension
• Creates a Python virtual environment
• Installs all required dependencies
• Configures Claude Desktop settings

Option 2: Manual Installation from DXT

# Install DXT CLI (if not already installed)
npm install -g @anthropic-ai/dxt

# Download the DXT package
wget https://github.com/doobidoo/MCP-Context-Provider/raw/main/mcp-context-provider-1.1.0.dxt

# Unpack the extension to your desired location
dxt unpack mcp-context-provider-1.1.0.dxt ~/mcp-context-provider

# Navigate to the installation directory
cd ~/mcp-context-provider

# Create and activate a Python virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install mcp>=1.9.4

Option 3: Installation from Source

# Clone the repository
git clone https://github.com/doobidoo/MCP-Context-Provider.git
cd MCP-Context-Provider

# Create and activate a Python virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

2. Configuration

Update your Claude Desktop configuration file:

Configuration File Location:

• Linux/Mac: ~/.config/claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

For Virtual Environment Installation (Recommended):

{
  "mcpServers": {
    "context-provider": {
      "command": "/path/to/mcp-context-provider/venv/bin/python",
      "args": ["/path/to/mcp-context-provider/context_provider_server.py"],
      "env": {
        "CONTEXT_CONFIG_DIR": "/path/to/mcp-context-provider/contexts",
        "AUTO_LOAD_CONTEXTS": "true"
      }
    }
  }
}

For System Python Installation:

{
  "mcpServers": {
    "context-provider": {
      "command": "python",
      "args": ["context_provider_server.py"],
      "cwd": "/path/to/MCP-Context-Provider",
      "env": {
        "CONTEXT_CONFIG_DIR": "./contexts",
        "AUTO_LOAD_CONTEXTS": "true"
      }
    }
  }
}

Important: Replace /path/to/mcp-context-provider with the actual installation path.

3. Verify Installation

Run the verification script to ensure everything is configured correctly:

python verify_install.py

4. Restart Claude Desktop

After updating the configuration, restart Claude Desktop to load the MCP server.

How It Works

Architecture

1. Context Provider Server: Python MCP server that loads JSON context files
2. Context Files: Tool-specific rules stored in /contexts directory
3. Claude Desktop Integration: MCP server registered in configuration
4. Automatic Loading: Context is injected at startup and persists across chats

Context Flow

Startup → Load Context Files → Register MCP Tools → Context Available in All Chats

Available Tools

Once loaded, the following tools are available in all chat sessions:

• get_tool_context: Get context rules for specific tool
• get_syntax_rules: Get syntax conversion rules
• list_available_contexts: List all loaded context categories
• apply_auto_corrections: Apply automatic syntax corrections

Context Files

The server loads context files from the /contexts directory:

• dokuwiki_context.json: DokuWiki syntax rules and preferences
• terraform_context.json: Terraform naming conventions and best practices
• azure_context.json: Azure resource naming and compliance rules
• git_context.json: Git commit conventions and workflow patterns
• general_preferences.json: Cross-tool preferences and standards

Context File Structure

Each context file follows this pattern:

{
  "tool_category": "toolname",
  "description": "Tool-specific context rules",
  "auto_convert": true,
  "syntax_rules": {
    "format_rules": "conversion patterns"
  },
  "preferences": {
    "user_preferences": "settings"
  },
  "auto_corrections": {
    "regex_patterns": "automatic fixes"
  },
  "metadata": {
    "version": "1.0.0",
    "applies_to_tools": ["tool:*"]
  }
}

Examples

DokuWiki Syntax Conversion

Input (Markdown):

# My Header
This is `inline code` and here's a [link](http://example.com).

Auto-converted to DokuWiki:

====== My Header ======
This is ''inline code'' and here's a [[http://example.com|link]].

Azure Resource Naming

Input: storage_account_logs_prod
Auto-corrected to: stlogsprod (following Azure naming conventions)

Git Commit Messages

Input: Fixed the login bug
Auto-corrected to: fix: resolve login authentication issue

Adding New Context

To add support for a new tool:

1. Create a new JSON file: contexts/{toolname}_context.json
2. Follow the standard context structure
3. Restart Claude Desktop to load the new context

The server automatically detects and loads any *_context.json files in the contexts directory.

Benefits

For Developers

• No need to re-establish context in new chats
• Automatic syntax corrections save time
• Consistent formatting across all work
• Best practices automatically applied

For Teams

• Shared context rules across team members
• Version-controlled standards
• Consistent code and documentation formatting
• Enterprise compliance automatically enforced

For Organizations

• Centralized context management
• Scalable across multiple tools
• Audit trail of context changes
• Easy deployment and updates

Advanced Usage

Custom Context Rules

Create your own context files by following the established pattern. The server supports:

• Regex-based auto-corrections
• Tool-specific preferences
• Conditional formatting rules
• Multi-tool context inheritance

Environment-Specific Context

Use environment variables to load different context sets:

{
  "env": {
    "CONTEXT_CONFIG_DIR": "./contexts/production",
    "ENVIRONMENT": "prod"
  }
}

Troubleshooting

Common Issues

1. Context not loading: Check file path in Claude Desktop config
2. Server not starting: Verify Python dependencies installed
3. Rules not applying: Check JSON syntax in context files

See TROUBLESHOOTING.md for detailed solutions.

Documentation

• Context Guide: Complete context file reference
• Developer Guide: Creating custom contexts
• Examples: Real-world usage examples
• Troubleshooting: Common issues and solutions

DXT Package Distribution

The MCP Context Provider is available as a Desktop Extension (DXT) package for easy distribution and installation:

• Package: mcp-context-provider-1.0.0.dxt (18.6 MB)
• Contents: Complete server with all dependencies bundled
• Platform: Windows, macOS, Linux with Python 3.8+
• Dependencies: Self-contained (no external pip requirements)

Building DXT Package

To build your own DXT package from source:

# Install DXT CLI
npm install -g @anthropic-ai/dxt

# Build the package
cd dxt
dxt pack

# The package will be created as mcp-context-provider-1.0.0.dxt

Distribution Notes

• The DXT package includes all Python dependencies (MCP SDK, Pydantic, etc.)
• Total unpacked size: ~45 MB including all dependencies
• Optimized for offline installation and deployment
• Compatible with corporate environments and air-gapped systems

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature/new-context
3. Add your context file to /contexts
4. Test with your Claude Desktop setup
5. Submit a pull request

License

MIT License - see LICENSE file for details.