AI Agent Template MCP Server

MCP Context Manager

The definitive MCP (Model Context Protocol) server for perfect AI-assisted development. This server transforms AI agents into expert developers that write flawless, secure, and well-tested code with zero hallucinations.

npm: https://www.npmjs.com/package/mcp-context-manager
GitHub: https://github.com/bswa006/mcp-context-manager

🚀 Overview

This MCP server is the missing piece for AI-assisted development, providing:

• 🧠 Zero Hallucinations: Context7 integration + multi-layer verification
• 📈 53% Better Code Quality: Enforced patterns + automated validation
• 🛡️ Security-First: Real-time vulnerability scanning
• 🧪 80%+ Test Coverage: Intelligent test generation
• ⚡ 30% Less Tokens: Efficient context management
• 🎯 Perfect Pattern Matching: Code indistinguishable from senior developers

🎉 What's New in v2.0.0

Complete UX Enhancement Suite

• Deep Codebase Analysis: Comprehensive pattern detection and architecture understanding
• Conversation Starters: Help AI understand your project instantly
• Token Optimization: 3-tier context system saving 70-95% tokens
• IDE Integrations: Auto-loading configs for Cursor, VS Code, and IntelliJ
• Persistence Automation: Git hooks, cron jobs, and monitoring
• Team Workflows: Onboarding, maintenance, and quality checklists
• One-Command Setup: Complete workflow from analysis to automation

🌟 Key Features

1. Agent Memory System

• Persistent Learning: Agents remember patterns, mistakes, and successes
• Context Awareness: Real-time tracking of current development session
• Performance Metrics: Continuous improvement through measurement

2. Hallucination Prevention

• API Verification: Every import and method checked before use
• Context7 Integration: Real-time documentation for latest APIs
• Pattern Validation: Ensures code matches existing conventions

3. Intelligent Code Generation

• Pattern Detection: Analyzes codebase to match style
• Security Scanning: Catches vulnerabilities before they happen
• Test Generation: Automatically creates tests for 80%+ coverage

4. Workflow Automation

• Guided Workflows: Step-by-step guidance for common tasks
• Proactive Prompts: AI guides itself through best practices
• Performance Tracking: Metrics for continuous improvement

🚀 Quick Start

Option 1: Use the Published npm Package (Recommended)

# Install globally
npm install -g mcp-context-manager

# Or use directly with npx
npx mcp-context-manager

Then add to your Claude Desktop config:

{
  "mcpServers": {
    "context-manager": {
      "command": "npx",
      "args": ["mcp-context-manager"]
    }
  }
}

Note: After updating Claude Desktop config, restart Claude Desktop completely for changes to take effect.

If you still see "0 tools enabled", try this alternative configuration:

{
  "mcpServers": {
    "context-manager": {
      "command": "node",
      "args": ["/path/to/global/node_modules/mcp-context-manager/dist/cli.js"]
    }
  }
}

To find the global node_modules path, run: npm root -g

Option 2: Clone and Build Locally

# Clone the repository
git clone https://github.com/bswa006/mcp-context-manager
cd mcp-context-manager

# Install dependencies
npm install

# Build the server
npm run build

Configuration

Claude Desktop

Add to your Claude Desktop configuration (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "context-manager": {
      "command": "node",
      "args": ["/path/to/ai-agent-template-mcp/dist/server.js"]
    }
  }
}

Cursor

Add to your Cursor settings:

{
  "mcp.servers": {
    "context-manager": {
      "command": "node",
      "args": ["/path/to/ai-agent-template-mcp/dist/server.js"]
    }
  }
}

Available Resources (AI Agent Self-Guidance)

Core Resources

• template://ai-constraints - CRITICAL rules AI must follow when generating code
• template://current-patterns - REQUIRED patterns to match in new code
• template://hallucination-prevention - Common AI mistakes and prevention guide
• template://naming-conventions - MANDATORY naming patterns to follow
• template://security-requirements - CRITICAL security rules (non-negotiable)
• template://api-signatures - Valid API methods to prevent hallucinations
• template://error-handling - REQUIRED error handling patterns

Agent Intelligence Resources

• template://agent-memory - Persistent memory of patterns and learnings
• template://agent-context - Real-time context for current session
• template://pattern-library - Comprehensive code patterns for all scenarios
• template://workflow-templates - Step-by-step guides for common tasks
• template://test-patterns - Testing strategies for 80%+ coverage

📚 Complete Tool Reference

Here's a comprehensive list of all 15 tools available in the MCP server:

Core Validation Tools

Tool	Purpose	Key Features
check_before_suggesting	Prevent hallucinations	Verifies imports, methods, and patterns exist before AI suggests code
validate_generated_code	Validate AI output	Checks generated code against project patterns and conventions
get_pattern_for_task	Pattern guidance	Provides exact patterns to follow for components, hooks, services, etc.
check_security_compliance	Security validation	Scans code for vulnerabilities and security issues
detect_existing_patterns	Pattern detection	Analyzes existing codebase to match coding style

Workspace & Project Tools

Tool	Purpose	Key Features
initialize_agent_workspace	Project setup	Creates PROJECT-TEMPLATE.md, CODEBASE-CONTEXT.md, and context files
analyze_codebase_deeply	Deep analysis	Comprehensive pattern detection, architecture understanding
complete_setup_workflow	One-command setup	Runs all setup tools in sequence for complete configuration

Testing & Performance Tools

Tool	Purpose	Key Features
generate_tests_for_coverage	Test generation	Creates tests to achieve 80%+ coverage with edge cases
track_agent_performance	Metrics tracking	Monitors token usage, validation scores, and improvements

UX Enhancement Tools (v2.0.0)

Tool	Purpose	Key Features
create_conversation_starters	AI context helper	Quick tasks, recent work, project overview for faster AI understanding
create_token_optimizer	Token savings	3-tier context system (minimal/standard/comprehensive) with ROI tracking
create_ide_configs	IDE integration	Auto-loading configs for Cursor, VS Code, IntelliJ
setup_persistence_automation	Auto-updates	Git hooks, cron jobs, monitoring, validation scripts
create_maintenance_workflows	Team collaboration	Onboarding guides, checklists, metrics dashboards, training materials

Available Tools (AI Self-Validation)

1. check_before_suggesting 🛑

CRITICAL: AI must use this before suggesting any code to prevent hallucinations.

{
  imports: string[];        // List of imports to verify
  methods: string[];        // List of methods/APIs to verify
  patterns?: string[];      // Code patterns to verify
}

2. validate_generated_code ✅

AI must validate all generated code against project patterns.

{
  code: string;            // Generated code to validate
  context: string;         // What the code is supposed to do
  targetFile?: string;     // Where this code will be placed
}

3. get_pattern_for_task 📋

Get the exact pattern to follow for a specific task.

{
  taskType: 'component' | 'hook' | 'service' | 'api' | 'test' | 'error-handling';
  requirements?: string[]; // Specific requirements
}

4. check_security_compliance 🔒

Verify code meets security requirements before suggesting.

{
  code: string;                    // Code to check
  sensitiveOperations?: string[];  // List of sensitive ops
}

5. detect_existing_patterns 🔍

Analyze existing code to match patterns when generating new code.

{
  directory: string;       // Directory to analyze
  fileType: string;        // Type of files to analyze
}

6. initialize_agent_workspace 🚀

Initialize complete AI agent workspace with templates and context.

{
  projectPath: string;     // Path to project
  projectName: string;     // Name of project
  techStack?: {           // Optional tech stack
    language?: string;
    framework?: string;
    uiLibrary?: string;
    testFramework?: string;
  };
}

7. generate_tests_for_coverage 🧪

Generate intelligent tests to achieve 80%+ coverage.

{
  targetFile: string;              // File to test
  testFramework?: string;          // jest, vitest, mocha
  coverageTarget?: number;         // Default: 80
  includeEdgeCases?: boolean;      // Include edge cases
  includeAccessibility?: boolean;  // Include a11y tests
}

8. track_agent_performance 📊

Track and analyze AI agent performance metrics.

{
  featureName: string;    // Feature completed
  timestamp: string;      // ISO timestamp
  metrics: {
    tokensUsed: number;
    timeElapsed: number;
    validationScore: number;
    securityScore: number;
    testCoverage: number;
    // ... more metrics
  };
}

9. analyze_codebase_deeply 🔬

Perform comprehensive analysis of codebase to understand patterns and architecture.

{
  projectPath: string;            // Path to analyze
  maxDepth?: number;             // Max directory depth (default: 5)
  excludePatterns?: string[];    // Patterns to exclude
}

10. create_conversation_sta

---

README truncated. View full README on GitHub.