MCP Manager

Modern web interface and CLI tool for managing MCP (Model Context Protocol) server configurations.

The web interface features a beautiful modern design with JetBrains Mono font, providing a clean and professional look:

Browse and install from 35+ pre-configured MCP servers with smart categorization and visual status indicators:

Organize servers by functionality for easier navigation and discovery:

Manage your installed servers with detailed configuration information:

Centralized environment variable management with .env file integration:

Edit.mcp.jsonfiles directly in the browser with syntax highlighting:

- One-click Installation: Simply click on any server card to install instantly
- Category Organization: Servers grouped by functionality (AI/ML, Databases, Cloud Services, etc.)
- Smart Search: Real-time filtering by name, description, or category
- Status Indicators: Visual dots showing installation status (green = installed, gray = not installed)

Special "S1-S14" servers install with zero configuration:

- S1-S7: Core utilities (Everything, Fetch, Filesystem, Git, Memory, Sequential Thinking, Time)
- S8-S11: Development tools (Playwright, Context7, Task Master, Xcode Build)
- S12-S14: Specialized tools (NixOS, shadcn/ui, Just Prompt)

- Visual Config Editor: Edit.mcp.jsondirectly in browser with syntax highlighting
- Environment Variables: Manage variables with .env file integration and "Fetch" buttons
- Current Servers: View, edit, and remove installed servers with intuitive controls

- Mandatory Updates: Automatic version checking with enforced updates
- Progress Tracking: Visual feedback during update process
- Seamless Experience: Auto-reload after successful updates

# Install globally npm install -g @oglabs/mcp # Launch web interface @oglabs/mcp --web

The interface automatically opens athttp://localhost:3333with full functionality.

Tech Stack Detection & MCP Server Recommendations

Transform your package into an intelligent MCP server that automatically detects project technology stacks and recommends the best MCP servers for your project:

The MCP server provides 4 powerful tools for AI agents:
- recommend-mcp-servers- Suggests optimal MCP servers based on provided tech stack
- search-mcp-servers- Search through 52+ available MCP servers
- get-mcp-server-info- Get detailed information about specific servers
- install-mcp-server- Install MCP servers directly to .mcp.json with confirmation workflow

The MCP server also provides intelligent prompts for guided tech stack analysis:
- analyze-tech-stack- Comprehensive project analysis with customizable depth
- discover-mcp-servers- Technology-specific MCP server discovery
- setup-project-mcp- Project setup assistant with team considerations
- quick-tech-assessment- Rapid tech stack assessment from minimal info
- confirm-installation- Interactive confirmation workflow for MCP server installation

# Start server with MCP Inspector for interactive testing npm run debug-mcp # Opens browser at http://localhost:6274 # Provides visual interface to test all MCP tools

Add this configuration to your Claude Desktopmcp.json:

{ "mcpServers": { "tech-stack-recommender": { "command": "npx", "args": ["-y", "@oglabs/mcp", "--server"], "env": {} } } }

- Prompt-Guided Tech Stack Analysis: Use intelligent prompts for comprehensive project analysis
- Smart MCP Recommendations: Get personalized server suggestions based on LLM-extracted keywords
- Server Discovery: Search and explore available MCP servers by technology
- Interactive Installation: Confirm and install servers one-by-one with environment validation
- Project Setup: Quickly identify what MCP servers would be most useful for any project

Launch the modern web interface for the best experience:

The web interface automatically opens in your default browser athttp://localhost:3333and provides:

- ⚡ Supafast Installation: 14 servers (S1-S14) can be installed instantly without configuration
- 🖱️ One-Click Install: Click directly on server cards to install servers with no required configuration
- 📦 Server Browser: Search and browse 26+ pre-configured MCP servers by category
- ⚙️ Current Servers: View, edit, and manage your configured servers
- ➕ Custom Servers: Add custom MCP server configurations with guided forms
- 📝 Direct Config Editing: Edit.mcp.jsonfiles directly in the browser
- 🔍 Advanced Search: Filter servers by name, category, or description
- 📱 Responsive Design: Works on desktop, tablet, and mobile devices
- 🎨 Modern UI: Clean design with JetBrains Mono font throughout

- S1-S7: Core tools (Everything, Fetch, Filesystem, Git, Memory, Sequential Thinking, Time)
- S8-S11: Development tools (Playwright, Context7, Task Master, Xcode Build)
- S12-S14: Specialized tools (NixOS, shadcn/ui, Just Prompt)

- GET /api/config- Retrieve current MCP configuration
- POST /api/config- Save MCP configuration
- GET /api/servers- Get pre-configured servers database
- GET /api/variables- Get saved environment variables
- POST /api/variables- Save environment variables
- GET /api/env-variables- Get .env file variables
- GET /api/version-check- Check for available updates
- POST /api/auto-update- Perform automatic update

Browse Servers → Click Card → Instant Install → Status Updates → Ready to Use

Current Servers Tab → Edit Server → Update Config → Auto-Save → Live Reload

Variables Tab → Add Variables → Fetch from .env → Auto-Fill → Save

Version Check → Update Available → Mandatory Modal → Progress Bar → Auto-Reload

- 🧠Prompt-Guided Analysis: Use intelligent prompts for comprehensive tech stack analysis
- 🎯Smart Recommendations: AI-powered MCP server suggestions based on LLM-extracted keywords
- 🔍Server Discovery: Search through 52+ available MCP servers by technology
- 📋Interactive Installation: Confirm and install servers one-by-one with validation
- 🔧4 MCP Tools: recommend-mcp-servers, search-mcp-servers, get-mcp-server-info, install-mcp-server
- 📝5 Smart Prompts: Guided analysis, discovery, setup, assessment, and installation confirmation
- 🧪MCP Inspector Support: Interactive testing with visual interface
- 📦Claude Desktop Ready: Easy integration with Claude Desktop MCP configuration
- 🚀LLM-Driven Workflow: Prompts first, then tools for optimal recommendations

- 🎨Modern Design: Beautiful depth-based UI with soft shadows and JetBrains Mono font
- 📱Fully Responsive: Optimized for desktop, tablet, and mobile devices
- 🖱️Interactive Server Cards: Click any card to install, hover for animations
- ⚡Instant Installation: Supafast servers (S1-S14) install with zero configuration
- 🎯Smart Status Indicators: Color-coded dots showing installation status
- 🔍Real-time Search: Filter 26+ servers by name, category, or description
- 📊Visual Dashboard: Manage installed servers with intuitive controls
- 📝Live Config Editor: Edit.mcp.jsondirectly with syntax highlighting
- 🔗Environment Integration: .env file support with convenient fetch buttons
- 🚀Auto-Update System: Mandatory updates with progress tracking
- 🎨Smooth Animations: Card interactions with CSS transitions and transforms

- 📋 Interactive menu-driven interface
- ⚡Supafast installation(S1-S14 servers with zero configuration)
- 🔧 35+ pre-configured MCP servers
- ➕ Add custom MCP servers
- ✏️ Edit existing servers
- ❌ Remove servers
- 📄 View current configuration
- 🔍 Search MCP servers
- 🎯 Automatic.mcp.jsonfile management

MCP Manager comes with 35+ verified pre-configured MCP servers including:

- 🗄️ Databases: Supabase, PostgreSQL, Redis, SQLite
- 🛠️ Development Tools: Git, Playwright, Xcode Build, PR Reviewer, NixOS
- ☁️ Cloud Services: AWS, Google Drive Enhanced, SharePoint
- 📁 File Systems: Filesystem, Memory, Fetch
- 🤖 AI/ML: Everything, Sequential Thinking, Context7, Just Prompt, GPT Researcher
- 🎯 Productivity: Task Master, ClickUp, shadcn/ui
- 🎵 Entertainment: Spotify
- 🔐 Authentication: Clerk
- 🚀 Automation: Skyvern
- 📚 Knowledge: Obsidian

MCP Manager manages.mcp.jsonfiles in your current working directory. The configuration format is:

{ "mcpServers": { "serverName": { "command": "npx", "args": ["-y", "@package/name"], "env": {"KEY": "value"} } } }

Installing a Supafast Server (Zero Config)

- Launch web interface:@oglabs/mcp --web - Browse Serverstab → Find any server markedS1-S14 - Single clickthe server card → Instant installation ⚡ - Green status dotappears → Server ready to use

Installing a Regular Server (With Config)

- Browse Serverstab → Find desired server (e.g., "Supabase") - ClickConfigurebutton → Configuration modal opens - Fill inrequired environment variables(DATABASE_URL, API_KEY, etc.) - ClickInstall→ Server installs with your configuration - Current Serverstab → Manage your new server - Variablestab → View all saved variables - Fetch from .envbutton → Auto-populate from local .env file - Add/Edit variables→ Set API keys, URLs, tokens - Save→ Variables available across all server configurations - Current Serverstab → See all installed servers - Editbutton → Modify server configuration in-browser - Raw Configtab → Direct.mcp.jsonediting with syntax highlighting - Save→ Changes applied immediately - Automatic checkon app load → Version comparison with npm - Mandatory update modal→ Cannot be closed if update required - "Update Now"button → Progress bar shows update status - Auto-reload→ Latest version loads automatically - Add to Claude Desktop configuration(~/.config/claude-desktop/mcp.json):

{ "mcpServers": { "tech-stack-recommender": { "command": "npx", "args": ["-y", "@oglabs/mcp", "--server"], "env": {} } } }

-

Restart Claude Desktopto load the new server

User: "Analyze my current project and recommend MCP servers" Claude: [uses analyze-tech-stack prompt to guide user analysis] → [user provides tech stack] → [uses recommend-mcp-servers tool with LLM-extracted keywords] Result: Personalized recommendations based on prompt-guided analysis User: "I have a React, TypeScript, PostgreSQL project. What MCP servers should I use?" Claude: [uses recommend-mcp-servers with keywords: languages=["TypeScript"], frameworks=["React"], databases=["PostgreSQL"]] Result: Curated recommendations for React/TypeScript/PostgreSQL development User: "Find MCP servers for working with databases" Claude: [uses search-mcp-servers with query "database"] Result: List of database-related MCP servers (PostgreSQL, MongoDB, SQLite, etc.) User: "What does the postgres MCP server do?" Claude: [uses get-mcp-server-info with serverKey "postgres"] Result: Detailed information about PostgreSQL MCP server User: "Install the recommended servers for me" Claude: [uses confirm-installation prompt with recommended servers] → [guides through confirmation] User: "Yes, install the filesystem and git servers" Claude: [uses install-mcp-server tool with confirmed: true for each server] Result: Servers installed to .mcp.json with success confirmation Recommended Workflow (Prompt-First Approach): 1. Start with analyze-tech-stack prompt for guided analysis 2. LLM extracts keywords from user's project description 3. Use recommend-mcp-servers tool with extracted keywords 4. Use confirm-installation prompt for user approval 5. Install servers one-by-one with install-mcp-server tool User: "Help me analyze my React project thoroughly" Claude: [uses analyze-tech-stack prompt with comprehensive depth] Result: Structured tech stack analysis with MCP recommendations User: "I need MCP servers for Python development" Claude: [uses discover-mcp-servers prompt with technology "Python"] Result: Curated list of Python-specific MCP servers with setup instructions

- recommend-mcp-servers: Provide tech stack data and get server recommendations
- search-mcp-servers: Try queries like "ai", "database", "github"
- get-mcp-server-info: Explore details of specific servers
- install-mcp-server: Test the installation workflow with confirmation steps

- analyze-tech-stack: Get structured prompts for comprehensive project analysis
- discover-mcp-servers: Find prompts for technology-specific server discovery
- setup-project-mcp: Get guided project setup prompts with team considerations
- quick-tech-assessment: Rapid assessment prompts for minimal project info
- confirm-installation: Interactive confirmation workflow for server installation

# Start server (connects via stdio) @oglabs/mcp --server # Alternative: Use npm script npm run mcp-server

Example MCP Server Recommendation Output

{ "providedTechStack": { "languages": ["JavaScript", "TypeScript"], "frameworks": ["React", "Express.js"], "databases": ["PostgreSQL"], "tools": ["Git", "Docker"], "deployment": ["Docker"] }, "recommendations": [ { "name": "PostgreSQL MCP Server", "key": "postgres", "description": "Connect to PostgreSQL databases", "reason": "Recommended for PostgreSQL", "confidence": 0.9, "package": "@modelcontextprotocol/server-postgres", "installCommand": "npx -y @modelcontextprotocol/server-postgres" }, { "name": "GitHub MCP Server", "key": "github", "description": "Interact with GitHub repositories", "reason": "Recommended for Git", "confidence": 0.8, "package": "@modelcontextprotocol/server-github", "installCommand": "npx -y @modelcontextprotocol/server-github" } ] }

Prompt: analyze-tech-stack Parameters: { "projectPath": ".", "analysisDepth": "comprehensive", "includeRecommendations": true } Result: Detailed structured analysis with MCP server recommendations

Prompt: discover-mcp-servers Parameters: { "technology": "React", "projectType": "web", "includeSetup": true } Result: Curated React-specific MCP servers with setup instructions

Prompt: setup-project-mcp Parameters: { "projectDescription": "E-commerce platform with React frontend and Node.js backend", "teamSize": "small", "experience": "intermediate", "priorities": ["productivity", "testing", "deployment"] } Result: Phased MCP server implementation plan with team considerations

Prompt: quick-tech-assessment Parameters: { "indicators": "package.json, Dockerfile, .github/workflows, cypress.json", "purpose": "mcp-recommendations" } Result: Rapid tech stack inference with immediate MCP opportunities

- Run@oglabs/mcp
- Select "Add a pre-configured MCP server"
- Choose from the list (e.g., "PostgreSQL")
- Follow the prompts for any required configuration
- Run@oglabs/mcp
- Select "Add a custom MCP server"
- Enter the server name, command, and arguments
- Optionally add environment variables
- Run@oglabs/mcp
- Select "Remove an MCP server"
- Choose the server to remove

For users who prefer command-line interfaces, MCP Manager provides a comprehensive CLI experience:

- Interactive Menu: Navigate through options with keyboard
- Server Management: Add, remove, and configure servers
- Search Functionality: Find servers by name or category
- Configuration Editing: Modify settings through prompts
- Cross-Platform: Works on Windows, macOS, and Linux

# MCP Server Mode npm run mcp-server # Start MCP server directly npm run debug-mcp # Start MCP server with inspector # Star Data Management npm run fetch-stars # Fetch GitHub stars for all servers npm run update-stale-stars # Update servers with stale star data npm run update-stars-7d # Update stars older than 7 days npm run update-stars-14d # Update stars older than 14 days

# Install globally and use npm install -g @oglabs/mcp # Different modes @oglabs/mcp # CLI mode @oglabs/mcp --web # Web interface @oglabs/mcp --server # MCP server mode # Local development npx @oglabs/mcp --server # Local MCP server npx @oglabs/mcp --web # Local web interface

# Clone the repository git clone https://github.com/kmhari/mcp-simple.git cd mcp-simple # Install dependencies npm install # Available development commands node mcp-manager.js # CLI mode node mcp-manager.js --web # Web mode node mcp-manager.js --server # MCP server mode npm run debug-mcp # MCP server with inspector

# Start MCP Inspector for development npm run debug-mcp # In another terminal, test specific tools echo '{"tool": "detect-tech-stack", "projectPath": "."}' | node mcp-server.js

Contributions are welcome! Please feel free to submit a Pull Request.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

Paid remote MCP for CLI tool MCP, structured receipts, usage logs, and audit-ready evidence for agent and CI workflows.

A flexible MCP server that executes arbitrary command-line tools defined in a configuration file.

Clean and audit messy MCP setups: stale servers, unused tools, context-heavy MCPs, outdated packages, and safe repair plans.

An MCP Server for the GitHub Actions API, enabling AI assistants to manage and operate GitHub Actions workflows.

Read-first Jenkins MCP server in Go for agent-driven build debugging. 20 tools including compare_builds, flaky-test detection, JUnit/Ginkgo failure parsing, and disk-cached console logs with on-disk path handoff. Write tools (trigger/stop/cancel) gated by JENKINS_MCP_READONLY env var.

A Python implementation of a Model Context Protocol server for interacting with Kestra.

Execute make targets from any Makefile in a safe and controlled environment.

Exposes Makefile targets as callable tools for AI assistants.