MCP Context Provider

MCP Context Provider

Static MCP server delivering persistent tool-specific context and automatic syntax rules.

19
Stars
4
Forks
5
Releases

Overview

An MCP Context Provider is a static MCP server that loads tool-specific context files from the /contexts directory and injects them into Claude Desktop at startup to prevent context loss across chat sessions. It serves as a central, persistent neural core, preserving per-tool rules, syntax preferences, and best practices, so conversations stay consistent across restarts. The server automatically discovers and loads context files (e.g., dokuwiki_context.json, terraform_context.json, azure_context.json, git_context.json, general_preferences.json) and exposes tool-specific rules and syntax mappings to all chats. It provides a suite of tools for interacting with and evolving context: core tools such as get_tool_context, get_syntax_rules, list_available_contexts, apply_auto_corrections; session management tools like execute_session_initialization and get_session_status; dynamic context management tools to create and update context files and add context patterns; and intelligent learning tools to analyze context effectiveness, suggest optimizations, provide proactive suggestions, and auto-optimize contexts. In phase 3, the system integrates with mcp-memory-service for persistent learning, memory-driven insights, and team knowledge propagation. The MCP Context Provider is available as a distributable DXT package or installable from source, designed for scalable, enterprise deployments with environment- and version-controlled configuration.

Details

Owner
doobidoo
Language
Python
License
Apache License 2.0
Updated
2026-02-14

Features

Persistent Context Across Sessions

Rules and preferences survive Claude Desktop restarts, eliminating the need to re-establish context in new chats.

Automatic Context Injection at Startup

Context is automatically loaded and injected into every conversation at startup.

Tool-Specific Context Rules and Syntax Preferences

Each tool receives dedicated context rules and syntax preferences for specialized guidance.

Auto-Corrections and Syntax Transformations

Automatic syntax corrections and transformations (e.g., Markdown to DokuWiki) for consistency.

Scalable and Extensible Tool Support

Easy to add new tools and context rules; scalable context network.

Central Neural Core Metaphor

A unified persistent core with neural pathways for tool-specific context rules.

Context Files Stored in /contexts

Loads predefined context files (e.g., dokuwiki_context.json, terraform_context.json, azure_context.json, git_context.json, general_preferences.json).

Intelligent Learning System with Memory Integration (Phase 3)

Phase 3 adds memory-service integration for persistent learning, optimization suggestions, and proactive context improvements.

Tags

MCPContextProviderpersistent-contexttool-specific-contextsyntax-rulesauto-correctionsClaude-Desktopmemory-serviceDXT-packageenterprise-ready