Skip to main content
Glama
MCP Context Master

MCP Context Master

by onigetoc

MCP Context Master

An intelligent Model Context Protocol (MCP) server that revolutionizes project bootstrapping by automatically discovering and downloading the most relevant contextual documentation for your development projects.

Overview

Context Master transforms how developers start new projects by providing intelligent context gathering through:

  1. Smart Project Analysis - Automatically analyzes your project dependencies

  2. Intelligent Priority System - Advanced scoring to recommend only the most valuable documentation

  3. Automated Documentation - Downloads comprehensive project context as organized markdown files

  4. Project Intelligence - Understands your tech stack and suggests the best resources

Related MCP server: Context7 MCP

Key Features

  • 🧠 Intelligent Analysis: Automatically analyzes package.json, requirements.txt, and project structure

  • 🎯 Smart Prioritization: Advanced scoring system that identifies which libraries truly need documentation

  • πŸ” Enhanced GitHub Search: Multi-criteria search with relevance scoring and health indicators

  • πŸ“₯ Automated Downloads: Bulk download of prioritized documentation with smart naming

  • πŸ“Š Project Intelligence: Understands your tech stack and provides tailored recommendations

  • πŸš€ One-Command Setup: Complete project context setup with a single setup_project_context call

Structure

  • Context Files: Downloaded .md files containing comprehensive project documentation

  • Project Documentation: Additional documentation for understanding your project dependencies

Installation

Quick Start

# Clone the repository
git clone https://github.com/onigetoc/mcp-context-master.git
cd mcp-context-master

# Install dependencies
npm install

# Build the project
npm run build

Environment Setup

Create a .env file with your GitHub token:

GITHUB_TOKEN=ghp_your_token_here

Get your GitHub token from: https://github.com/settings/tokens

Configuration

MCP Client Configuration

Add to your MCP client configuration:

For Kiro IDE (.kiro/settings/mcp.json):

{
  "mcpServers": {
    "mcp-context-master": {
      "command": "node",
      "args": [
        "C:\\path\\to\\mcp-context-master\\build\\index.js"
      ],
      "env": {
        "GITHUB_TOKEN": "your_github_token_here"
      }
    }
  }
}

For Claude Desktop:

{
  "mcpServers": {
    "mcp-context-master": {
      "command": "node",
      "args": ["/path/to/mcp-context-master/build/index.js"],
      "env": {
        "GITHUB_TOKEN": "your_github_token_here"
      }
    }
  }
}

Available Tools

1. setup_project_context ⭐ Main Tool

Initialize and setup Context Master for a project. Analyzes your project dependencies and downloads documentation for important libraries.

2. add_project_context

Fetch and download context for a single library and add it to the project.

3. list_available_contexts

List all available context files in your project.

4. read_specific_context

Read the content of a specific context file.

5. read_template

Read a Context Master template file from your project.

6. search_github

Search GitHub repositories for project context.

7. coding_assistant

Detect your current coding assistant and determine the appropriate context to use.

Usage Examples

Initialize a New Project

To set up Context Master for your project:

User: "Initialize context master for my project"
Tool: setup_project_context will analyze your package.json/requirements.txt and download relevant documentation

Add Context for a Specific Library

To add documentation for a library:

User: "Add project context for remotion"
Tool: add_project_context will search and download remotion documentation

List Downloaded Context

To see what documentation you have:

User: "List available contexts"
Tool: list_available_contexts will show all downloaded context files

View Documentation

To read a specific context file:

User: "Read the react context"
Tool: read_specific_context will display the documentation

Requirements

  • Node.js 18+

  • GitHub token (for repository search)

  • Internet connection (for documentation downloads)

  • MCP client (Claude Desktop, VSCode, Kiro, etc.)

License

MIT

Available Tools

9 tools
add_project_contextA

Adds knowledge context for a library using EXACT package names from package.json or confirmed library names. If you're unsure about the exact library name, use the search workflow first: 1) Search GitHub for the library, 2) Confirm the correct repository, 3) Use this tool with the exact name. For example: Use 'remotion' (not '@remotion/captions'), 'react' (not 'react-dom'), '@tanstack/react-query' (not 'react-query'). If the user mentions a feature or topic (like 'captions', 'routing'), that should be used as a topic parameter, not as part of the library name. IMPORTANT: Always provide the absolute path to the user's project directory as projectPath parameter.

Inputs: libraryName (required), topic, tokens, projectPath.

initialize_context_masterA

🚨 MANDATORY STEP 1: Initialize Context Master. When user says 'init context master' or '/cm-init', call this tool FIRST, then IMMEDIATELY call setup_project_context. DO NOT ask user for confirmation between steps. DO NOT ask user to create files manually. This tool auto-creates all required files including cm-ai-infos.yaml. CRITICAL: After this tool completes, you MUST call setup_project_context with the same projectPath. No exceptions.

Inputs: projectPath.

list_added_contexts_detailsB

Lists all available knowledge files with metadata including title, description, and creation date.

Inputs: no named parameters.

list_available_contextsA

Lists all available knowledge files from the knowledge manifest.

Inputs: no named parameters.

read_specific_contextA

Reads the content of a specific knowledge file.

Inputs: fileName (required).

read_templateA

Read a Context Master template file from the user's project .context-master directory

Inputs: templateName (required), projectPath.

search_library_advisorA

Search for libraries and get guidance on the correct workflow for adding context. Use this when you're unsure about exact library names or when users mention features/topics. This tool will find the correct repository and suggest the proper Context7 workflow with appropriate topics. Example: search for 'Remotion' to find 'remotion-dev/remotion', then suggest using Context7 with topic 'captions' or 'srt'.

Inputs: query (required), topic.

setup_project_contextB

🚨 MANDATORY STEP 2: Complete Context Master setup. This tool is AUTOMATICALLY called after initialize_context_master. When user says 'init context master', you MUST call initialize_context_master first, then IMMEDIATELY call this tool. DO NOT ask user for confirmation. DO NOT ask user to create files manually. This tool auto-handles all missing configurations and completes the setup process.

Inputs: projectPath, maxDependencies.

update_agents_fileA

Update project's AGENTS.md file with Context Master instructions. Reads the template from .context-master directory and updates the project's AGENTS.md file. Automatically detects the current project directory.

Inputs: projectPath.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 9 tools

    Disambiguation3/5

    Two listing tools (list_available_contexts and list_added_contexts_details) have overlapping descriptions, both listing knowledge files, though one includes metadata. The setup tools (initialize_context_master and setup_project_context) are sequential but clearly defined. Other tools target distinct actions, so overall ambiguity is moderate.

    Naming Consistency4/5

    Tool names consistently use snake_case verb-first pattern (search, initialize, setup, add, list, read, update). Minor inconsistencies: 'search_library_advisor' uses an agent-like noun rather than a resource, and 'list_added_contexts_details' is verbose. But the pattern is predictable and readable.

    Tool Count5/5

    9 tools is a well-scoped number for a context management server, covering setup, searching, adding, listing, reading, and updating. No redundant or excessive tools, and each earns its place.

    Completeness3/5

    The toolset covers setup, add, list, read, and update agents file, but lacks delete/remove context operations and any update mechanism for existing context entries. This creates a notable gap in lifecycle management, though the core workflow of adding and reading context is covered.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Automatically provides AI assistants with contextual, version-specific documentation for Python project dependencies by scanning pyproject.toml files. Eliminates manual package lookup and enables more accurate coding assistance through seamless integration with AI tools.
      4
      45 PyPI
      1
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Provides LLMs with up-to-date, version-specific documentation and code examples directly from library sources, eliminating outdated training data and hallucinated APIs by fetching current documentation at prompt time.
      2
      MIT
    • A
      license
      A
      quality
      C
      maintenance
      Provides AI assistants with intelligent access to project documentation and API references through smart search, contextual rules, and Docset integration. Enables AI to understand project-specific conventions, patterns, and official framework documentation without token limits.
      11
      109 npm
      6
      MIT