roundtable-ai 0.3.0

Zero-configuration MCP server that unifies multiple AI coding assistants (Codex, Claude Code, Cursor, Gemini) through intelligent auto-discovery and standardized interface

Roundtable AI MCP Server

Connect to Multiple AI Coding Tools at Once - Use Codex, Claude, Cursor, and Gemini from one simple interface.

What You Get

🚀 Quick Setup - Install once, use everywhere

🔧 Works Out of the Box - Finds and sets up your AI tools automatically. No config files to edit.

🔗 Same Commands for Everything - Switch between AI tools without learning different commands

📦 Works with Popular Editors - Use it in VS Code, Claude Desktop, Cursor, and 20+ other coding tools

🛡️ Only Shows Working Tools - Remembers which AI tools you have installed and only shows the ones that actually work

Why You Need This

1. Stop Wrestling with Multiple AI Tools

The Problem: Each AI tool (Codex, Claude, Gemini) has different commands, setup steps, and ways to log in. This wastes time switching between them.

Our Solution: Roundtable finds your AI tools and lets you use them all the same way. One command, all your AI tools.

2. Use Any AI Tool the Same Way

The Problem: Each AI tool needs different integration code and handles errors differently.

Our Solution: All AI tools work exactly the same through Roundtable. Switch tools without changing your code.

3. Know Which Tools Actually Work

The Problem: You waste time trying to use AI tools that aren't installed or configured properly.

Our Solution: Roundtable remembers which AI tools work on your system and only shows you the ones that are actually installed.

Quick Start

# Install once, access all AI assistants
pip install roundtable-ai

# Check which AI tools are available
roundtable-ai --check

# Start unified AI assistant interface
roundtable-ai

# Use specific assistants only
roundtable-ai --agents codex,gemini

Supported AI Assistants

• Codex - Advanced code generation and assistance
• Claude Code - Anthropic's Claude with enhanced code capabilities
• Cursor - AI-powered code editor integration
• Gemini - Google's Gemini AI assistant

Only assistants that are actually installed and configured will be enabled automatically.

Zero-Configuration Intelligence

Automatic Discovery

• Runs --help commands to verify CLI tool availability
• Smart caching saves availability results to avoid repeated checks
• Environment-aware adaptation to different development environments

Priority-Based Configuration

1. Command line flags (--agents codex,gemini) - Highest priority
2. Environment variables (CLI_MCP_SUBAGENTS)
3. Availability cache (~/.roundtable/availability_check.json)
4. Defaults (all available tools) - Lowest priority

Installation & Usage

Multiple Installation Options

# Primary installation method
pip install roundtable-ai

# Multiple command aliases (all equivalent)
roundtable-ai                    # Short and convenient
roundtable-mcp-server           # Descriptive
python -m roundtable_mcp_server # Module execution

Check Available AI Tools

roundtable-ai --check

This command:

• Tests each CLI tool by running --help commands
• Saves results to ~/.roundtable/availability_check.json
• Shows detailed availability report

Start the Unified Interface

# Use all available AI assistants
roundtable-ai

# Use only specific assistants
roundtable-ai --agents codex,claude

# Override availability detection
CLI_MCP_IGNORE_AVAILABILITY=true roundtable-ai

MCP Client Integration

Claude Desktop Configuration

Add to your ~/.config/claude_desktop_config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Universal MCP Compatibility

Works with any MCP-compatible client including:

• Claude Desktop
• VS Code with MCP extensions
• Custom MCP applications
• Future MCP-compatible tools

IDE & Editor Integration

Roundtable AI MCP Server integrates with 26 different IDEs and AI coding tools. Below are the installation guides for each platform.

🎯 Primary IDEs (Most Popular)

1. Cursor - AI-First Code Editor

Create or edit .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Alternative using npx:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "npx",
      "args": ["-y", "@roundtable/mcp-server"],
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

2. Claude Code - Anthropic's Official CLI

Use Claude Code's built-in MCP management:

# Quick installation
claude mcp add roundtable-ai roundtable-ai

# With specific working directory
claude mcp add roundtable-ai roundtable-ai --env CLI_MCP_WORKING_DIR=/path/to/project

# Verify installation
claude mcp list

# Remove if needed
claude mcp remove roundtable-ai

Manual Configuration (~/.config/claude_desktop_config.json):

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

3. Windsurf - AI-Powered Development Environment

Edit ~/.codeium/windsurf/mcp_config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

4. VS Code - Microsoft's Editor

Add to your settings.json (Ctrl/Cmd + Shift + P → "Open Settings JSON"):

{
  "mcp.servers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

5. GitHub Copilot - GitHub's AI Coding Assistant

Add to your GitHub Copilot settings or workspace configuration:

{
  "github.copilot.mcp.servers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

---

🖥️ Desktop IDEs

JetBrains AI Assistant - IntelliJ, PyCharm, WebStorm, etc.

1. Settings Path: Settings > Tools > AI Assistant > Model Context Protocol (MCP)
2. Add New Server: Click "+" to add new MCP server
3. Configuration:

   {
     "name": "roundtable-ai",
     "command": "roundtable-ai",
     "transport": "stdio",
     "env": {
       "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
       "CLI_MCP_WORKING_DIR": "/path/to/your/project"
     }
   }
   

4. Apply & Restart: Apply settings and restart the IDE

Visual Studio 2022 - Microsoft's Flagship IDE

Create mcp_config.json in your project root:

{
  "servers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "transport": "stdio",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Alternative: Use Extensions > Manage Extensions > Search for "Roundtable AI"

Zed - High-Performance Code Editor

Add to settings.json (Cmd/Ctrl + ,):

{
  "context_servers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Extension Alternative: Search for "Roundtable AI" in Zed Extensions

---

💻 CLI Tools

Gemini CLI - Google's Gemini Command-Line Interface

Edit ~/.gemini/settings.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Rovo Dev CLI - Atlassian's Development CLI

Configure via rovo config command:

# Add MCP server
rovo mcp add roundtable-ai roundtable-ai

# Verify
rovo mcp list

Manual configuration in ~/.rovo/config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

Amazon Q Developer CLI - Amazon's AI Development Assistant

Edit configuration in ~/.aws/q-developer/config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Crush - Terminal-Based AI Assistant

Create or edit crush.json in your project:

{
  "mcp": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "transport": "stdio",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Warp - AI-Powered Terminal

Configure via Warp settings:

1. Open Settings: Cmd/Ctrl + ,
2. Navigate to: Features > AI > MCP Servers
3. Add Server:

   {
     "name": "roundtable-ai",
     "command": "roundtable-ai",
     "env": {
       "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
     }
   }
   

---

🤖 AI Assistants

Claude Desktop - Anthropic's Desktop Application

Edit ~/.config/claude_desktop_config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

Cline - AI Assistant Extension

One-Click Install:

1. Open Cline MCP Server Marketplace
2. Search for "Roundtable AI"
3. Click "Install"

Manual Configuration in cline_mcp_settings.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

BoltAI - AI Assistant Application

1. Open BoltAI Settings
2. Navigate to: Plugins > MCP Servers
3. Add New Server:
   • Name: roundtable-ai
   • Command: roundtable-ai
   • Environment Variables:

     CLI_MCP_SUBAGENTS=codex,claude,cursor,gemini
     CLI_MCP_WORKING_DIR=/path/to/your/project
     

Perplexity Desktop - AI Search and Research Assistant

Configure in Perplexity settings:

{
  "mcpConfig": {
    "servers": {
      "roundtable-ai": {
        "command": "roundtable-ai",
        "env": {
          "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
        }
      }
    }
  }
}

Qodo Gen - AI Code Generation and Analysis Tool

Add to Qodo Gen configuration:

{
  "mcp_servers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
        "CLI_MCP_WORKING_DIR": "/path/to/your/project"
      }
    }
  }
}

---

🛠️ Specialized Tools

Opencode - Open-Source AI Code Editor

Add to opencode_config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

OpenAI Codex - OpenAI's Code Generation Model Interface

Edit config.toml:

[mcp_servers.roundtable-ai]
command = "roundtable-ai"
env = { CLI_MCP_SUBAGENTS = "codex,claude,cursor,gemini" }

Kiro - AI Development Assistant

Configure in ~/.kiro/config.json:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

Trae - AI Development Environment

Add to Trae workspace configuration:

{
  "mcp": {
    "servers": {
      "roundtable-ai": {
        "command": "roundtable-ai",
        "env": {
          "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
          "CLI_MCP_WORKING_DIR": "/path/to/your/project"
        }
      }
    }
  }
}

LM Studio - Local Language Model Interface

One-Click Install:

1. Navigate to Program > Install > Edit mcp.json
2. Search for "Roundtable AI" in marketplace
3. Click "Install"

Manual Configuration:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

Zencoder - AI-Powered Coding Assistant

Configure via Zencoder settings panel:

{
  "mcp_configuration": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

Augment Code - AI-Powered Code Completion

Add to Augment Code workspace settings:

{
  "mcpServers": {
    "roundtable-ai": {
      "command": "roundtable-ai",
      "transport": "stdio",
      "env": {
        "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini"
      }
    }
  }
}

Roo Code - AI Development Environment

Configure in Roo Code project settings:

{
  "ai_assistants": {
    "mcp_servers": {
      "roundtable-ai": {
        "command": "roundtable-ai",
        "env": {
          "CLI_MCP_SUBAGENTS": "codex,claude,cursor,gemini",
          "CLI_MCP_WORKING_DIR": "/path/to/your/project"
        }
      }
    }
  }
}

---

🔧 Configuration Tips

Environment Variables Reference

# Specify which AI assistants to enable
CLI_MCP_SUBAGENTS="codex,claude,cursor,gemini"

# Set working directory for project context
CLI_MCP_WORKING_DIR="/path/to/your/project"

# Enable debug logging
CLI_MCP_DEBUG=true

# Override availability checking
CLI_MCP_IGNORE_AVAILABILITY=true

Command Alternatives

All IDEs support these equivalent commands:

• roundtable-ai (primary command)
• roundtable-mcp-server (descriptive alias)
• python -m roundtable_mcp_server (Python module)
• npx @roundtable/mcp-server (NPM package - coming soon)

Verification

After installation, verify the integration works:

# Check server availability
roundtable-ai --check

# Test connection (varies by IDE)
# Most IDEs will show "Roundtable AI" in their AI assistant panel

Troubleshooting

1. Server not found: Ensure roundtable-ai is in your PATH
2. Permission denied: Run chmod +x $(which roundtable-ai)
3. Config not loaded: Check file paths and JSON syntax
4. No AI tools detected: Run roundtable-ai --check first

Available MCP Tools

Once integrated, you get access to:

Availability Checks

• check_codex_availability - Verify Codex CLI status
• check_claude_availability - Verify Claude Code CLI status
• check_cursor_availability - Verify Cursor CLI status
• check_gemini_availability - Verify Gemini CLI status

Unified Task Execution

• execute_codex_task - Run coding tasks through Codex
• execute_claude_task - Run coding tasks through Claude Code
• execute_cursor_task - Run coding tasks through Cursor
• execute_gemini_task - Run coding tasks through Gemini

Advanced Configuration

Environment Variables

# Specify which assistants to enable
export CLI_MCP_SUBAGENTS="codex,gemini"

# Set project working directory
export CLI_MCP_WORKING_DIR="/path/to/project"

# Enable all tools regardless of availability
export CLI_MCP_IGNORE_AVAILABILITY=true

# Enable debug logging
export CLI_MCP_DEBUG=true

Command Line Flags

# Get comprehensive help
roundtable-ai --help

# Check availability and save results
roundtable-ai --check

# Start with specific agents (overrides all other settings)
roundtable-ai --agents cursor,claude

What Teams Get

Faster Development

• Quick Setup: Most teams set up new developers in minutes, not hours
• Less Debugging: Fewer "why isn't this AI tool working?" problems
• Easier Switching: Try different AI tools on the same problem without hassle
• Faster Onboarding: New team members get all AI tools working immediately

Same Setup Everywhere

• All developers use the same AI tools the same way
• No more "works on my machine" problems with AI tools
• One place to check all AI tool usage

Works with Future Tools

• Built on MCP standard that major companies support
• Switch AI providers easily without changing your setup
• No vendor lock-in - use any AI tools you want
• New AI tools will work automatically when they support MCP

Requirements

System Requirements

• Python 3.8+
• One or more AI CLI tools installed:
  • Codex CLI (for Codex support)
  • Claude Code CLI (for Claude support)
  • Cursor CLI (for Cursor support)
  • Gemini CLI (for Gemini support)

Dependencies

The server automatically handles all dependencies:

• FastMCP framework for high-performance MCP protocol implementation
• Async/await support for concurrent AI assistant requests
• Intelligent session management and error handling

How It Works

Built for Real Use

• FastMCP Framework - Fast, reliable MCP server that handles multiple requests
• Easy to Extend - Adding new AI tools is straightforward
• Uses Minimal Resources - Won't slow down your computer
• Fails Safely - If one AI tool breaks, the others keep working

Keeps Things Working

• Remembers your conversation when switching between AI tools
• Runs commands in the right folder for your project
• Uses your existing AI tool logins - no new passwords to remember
• Handles multiple requests at once without slowing down

Development & Testing

Testing the Installation

# Run the test suite
python -m roundtable_mcp_server.test_server

# Check server functionality
roundtable-ai --check

Building from Source

git clone https://github.com/askbudi/roundtable
cd roundtable_mcp_server
pip install -e .[dev]

Development Tools

Includes comprehensive development support:

• Black code formatting
• Ruff linting
• MyPy type checking
• Pytest test suite
• Pre-commit hooks

Troubleshooting

Common Issues

No AI tools detected?

• Run roundtable-ai --check to see detailed availability report
• Ensure CLI tools are properly installed and in your PATH
• Check that API keys are configured for tools that require them

MCP client not connecting?

• Verify MCP client configuration matches the command aliases
• Check that the working directory is accessible
• Enable debug logging with CLI_MCP_DEBUG=true

Want to force enable specific tools?

• Use CLI_MCP_IGNORE_AVAILABILITY=true to bypass availability checking
• Use --agents flag for precise control: roundtable-ai --agents codex,claude

Debug Information

• Debug logs are written to: roundtable_mcp_server.log
• Availability cache location: ~/.roundtable/availability_check.json
• Use --help for comprehensive command documentation

License

GNU Affero General Public License v3

About Roundtable AI

Roundtable AI develops intelligent infrastructure and tooling solutions to enhance AI-assisted development workflows. We focus on eliminating complexity and friction in developer tools while maintaining the highest standards of reliability and performance.

Learn more: askbudi.ai Repository: github.com/askbudi/roundtable Support: [e-mail]

---

Ready to unify your AI coding assistants? Install roundtable-ai and experience zero-configuration AI integration in under 5 minutes.