Pal

v1.0.0Coding Agentsstable

The power of Claude Code / GeminiCLI / CodexCLI + [Gemini / OpenAI / OpenRouter / Azure / Grok / Ollama / Custom Model / All Of The Above] working as one.

palmcpai-integration
Share:
11,553
Stars
0
Downloads
0
Weekly
0/5

What is Pal?

Pal is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to power of claude code / geminicli / codexcli + [gemini / openai / openrouter / azure / grok / ollama / custom model / all of the above] working as one.

The power of Claude Code / GeminiCLI / CodexCLI + [Gemini / OpenAI / OpenRouter / Azure / Grok / Ollama / Custom Model / All Of The Above] working as one.

This server falls under the Coding Agents category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • The power of Claude Code / GeminiCLI / CodexCLI + [Gemini /

Use Cases

Unify Claude Code, GeminiCLI, or CodexCLI with multiple AI models.
Switch between Gemini, OpenAI, OpenRouter, Azure, Grok, or Ollama seamlessly.
LicenseNOASSERTION
Languagepython
Versionv1.0.0
UpdatedMay 20, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx pal

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Pal

PAL MCP Server (Polyglot AI Layer) is a multi-model orchestration server that lets Claude Code, Gemini CLI, and Codex CLI delegate tasks to any combination of AI providers — Gemini, OpenAI, OpenRouter, Azure, Grok, Ollama, and more — through a unified MCP interface. It exposes purpose-built tools like thinkdeep for extended reasoning, consensus for getting opinions from multiple models simultaneously, codereview for professional code reviews, precommit for change validation, and debug for root cause analysis. Teams use PAL to route different coding tasks to the best-fit model without switching tools or managing separate API integrations.

Prerequisites

  • Python 3.10 or newer
  • Git installed on your system
  • uv package manager (install from https://docs.astral.sh/uv/)
  • At least one AI provider API key: GEMINI_API_KEY, OPENAI_API_KEY, or OPENROUTER_API_KEY
  • Claude Desktop, Claude Code, or another MCP-compatible client
1

Clone the repository

Clone the PAL MCP Server repository to your local machine.

git clone https://github.com/BeehiveInnovations/pal-mcp-server.git
cd pal-mcp-server
2

Set your API keys

Set at least one provider API key. You can configure multiple providers and PAL will route to the best available model, or you can specify a model explicitly in your prompts.

export GEMINI_API_KEY=your-gemini-key
export OPENAI_API_KEY=your-openai-key
export OPENROUTER_API_KEY=your-openrouter-key
# For Grok:
export X_API_KEY=your-xai-key
3

Run the automatic setup

Run the setup script to install dependencies and automatically configure Claude Desktop, Claude Code, Gemini CLI, Codex CLI, and Qwen CLI.

./run-server.sh
4

Or add manually to your MCP client config

Add the PAL server to your MCP client configuration file manually using uvx. Adjust DISABLED_TOOLS to expose only the tools you need.

{
  "mcpServers": {
    "pal": {
      "command": "bash",
      "args": ["-c", "uvx --from git+https://github.com/BeehiveInnovations/pal-mcp-server.git pal-mcp-server"],
      "env": {
        "GEMINI_API_KEY": "your-gemini-api-key",
        "OPENAI_API_KEY": "your-openai-api-key",
        "DISABLED_TOOLS": "analyze,refactor,testgen,secaudit,docgen,tracer",
        "DEFAULT_MODEL": "auto"
      }
    }
  }
}
5

Configure performance and tool settings

Optionally create a .env file in the project root to tune conversation behavior and enable or disable specific tools.

# .env
DEFAULT_MODEL=auto
LOG_LEVEL=INFO
CONVERSATION_TIMEOUT_HOURS=6
MAX_CONVERSATION_TURNS=50
# Enable all tools by leaving DISABLED_TOOLS empty or omitting it
DISABLED_TOOLS=
6

Test the server

Open your MCP client and ask PAL to perform a code review or brainstorm using a specific model. The server will route the request to the appropriate provider.

Pal Examples

Client configuration

Full MCP configuration with multiple provider keys and selective tool enabling via DISABLED_TOOLS.

{
  "mcpServers": {
    "pal": {
      "command": "bash",
      "args": ["-c", "uvx --from git+https://github.com/BeehiveInnovations/pal-mcp-server.git pal-mcp-server"],
      "env": {
        "GEMINI_API_KEY": "your-gemini-api-key",
        "OPENAI_API_KEY": "your-openai-api-key",
        "OPENROUTER_API_KEY": "your-openrouter-api-key",
        "DISABLED_TOOLS": "tracer,docgen",
        "DEFAULT_MODEL": "auto"
      }
    }
  }
}

Prompts to try

Example prompts demonstrating multi-model orchestration, code review, and planning capabilities.

- "Use pal to do a codereview with gemini-pro, then use planner to generate a fix plan"
- "Run thinkdeep on this algorithm and identify edge cases using o3"
- "Get consensus from multiple models on whether this API design is RESTful"
- "Use precommit to validate my staged changes before I commit"
- "Debug this error with pal — trace the root cause and suggest a fix using Gemini Flash"
- "Use clink with cli_name='gemini' role='planner' to draft a phased migration plan"

Troubleshooting Pal

Server starts but no tools appear in the MCP client

Check that DISABLED_TOOLS does not list the tools you expect. By default, analyze, refactor, testgen, secaudit, docgen, and tracer are disabled. The always-on tools are chat, thinkdeep, planner, consensus, codereview, precommit, debug, apilookup, challenge, and clink.

Model routing fails or returns 'no provider configured'

Ensure at least one API key environment variable is set (GEMINI_API_KEY, OPENAI_API_KEY, OPENROUTER_API_KEY, or X_API_KEY). For local models, ensure Ollama is running on localhost:11434. The DEFAULT_MODEL=auto setting picks the first available provider.

uvx command not found when starting server

Install uv with 'curl -LsSf https://astral.sh/uv/install.sh | sh' and add ~/.local/bin to your PATH. Alternatively, use the ./run-server.sh script which handles uv detection automatically.

Frequently Asked Questions about Pal

What is Pal?

Pal is a Model Context Protocol (MCP) server that power of claude code / geminicli / codexcli + [gemini / openai / openrouter / azure / grok / ollama / custom model / all of the above] working as one. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Pal?

Follow the installation instructions on the Pal GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Pal?

Pal works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Pal free to use?

Yes, Pal is open source and available under the NOASSERTION license. You can use it freely in both personal and commercial projects.

Browse More Coding Agents MCP Servers

Explore all coding agents servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "pal": { "command": "npx", "args": ["-y", "pal"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Pal?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides