Polydev

Provides multi-model perspectives via a single MCP server for coding tasks.
  • javascript

0

GitHub Stars

javascript

Language

4 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

Polydev MCP Server enables you to query multiple frontier models simultaneously from a single control plane. By routing your agent’s requests through Polydev, you gain diverse perspectives—across GPT-5, Claude, Gemini, and Grok—within one request to synthesize richer, less biased answers for coding tasks.

How to use

You connect your MCP client to Polydev MCP Server to request perspectives from multiple models in one go. Once connected, your agent can ask for refactoring guidance, architecture opinions, or model-specific insights, and Polydev combines the results into a single, synthesized response.

How to install

Prerequisites: you need Node.js 18+ and a functioning shell. If you prefer a hosted service, you can start immediately with the hosted option; if you want to self-host, follow the self-hosting steps.

# Option A: Quick start with hosted service

# Install the MCP server locally (hosted by Polydev)
npx polydev-ai@latest

# Set your authentication token (get it from your dashboard)
export POLYDEV_USER_TOKEN="pd_your_token_here"

# Then start using the MCP from your client as configured below

# Option B: Self-host with your own API keys

# Clone the project, install dependencies, and run
git clone https://github.com/polydev-ai/polydev.git
cd polydev && npm install

# Copy environment template and add keys
cp .env.example .env.local
# Add your API keys to .env.local

# Start the local MCP server
npm run dev

Additional notes

Token management: obtain your token from the designated dashboard and supply it to the MCP client configuration. Your CLI tools can be authenticated to use your existing subscriptions, so you don’t incur extra API costs.

Configuration at a glance: you can configure the MCP client to point to Polydev with the provided npx command and supply the token via environment variables as shown in the IDE integration details.

IDE configuration

Claude Code users can add Polydev as an MCP source with a single command, or embed the MCP server in their user configuration.

# Claude Code quick add
claude mcp add polydev -- npx -y polydev-ai@latest

# Or in the Claude JSON config (example)
{
  "mcpServers": {
    "polydev": {
      "command": "npx",
      "args": ["-y", "polydev-ai@latest"],
      "env": {
        "POLYDEV_USER_TOKEN": "pd_your_token_here"
      }
    }
  }
}

Cursor / Windsurf / Cline users can integrate Polydev by adding an MCP entry in their config files with the same command and environment variable.

{
  "mcpServers": {
    "polydev": {
      "command": "npx",
      "args": ["-y", "polydev-ai@latest"],
      "env": {
        "POLYDEV_USER_TOKEN": "pd_your_token_here"
      }
    }
  }
}

Usage patterns

From your agent prompt, you can request perspectives by invoking the MCP tool with a prompt. For example, you can ask for refactoring guidance, caching strategy advice, or multi-model opinions on architecture decisions.

{
  "tool": "get_perspectives",
  "arguments": {
    "prompt": "How should I refactor this authentication flow?"
  }
}

Available tools

get_perspectives

Requests multi-model perspectives for a given prompt and returns a synthesized answer.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational