PRD Studio MCP Server

Connect your AI coding agents (Cursor, Claude Code, etc.) to your PRD Studio architectural blueprints via the Model Context Protocol (MCP).

Features

• Project Context: Grant AI agents real-time access to your user stories, feature sets, and technical specs.
• Build Roadmap: Sync the implementation sequence directly into your agent's context.
• Safe Proxy: Securely forwards requests to the PRD Studio Cloud API without exposing database credentials.

Quick Start (Automatic)

Run via npx (requires an API key from your PRD Studio Profile):

# Basic (Requires clicking 'Sync to Agent' in UI)
npx -y prd-studio-mcp --key=YOUR_API_KEY

# Persistent (Locked to a specific project)
npx -y prd-studio-mcp --key=YOUR_API_KEY --project=PROJECT_ID

Manual Installation

If you prefer to run it from source:

1. Clone the main repository:

   git clone https://github.com/wisdomgreat/prd-studio.git

2. Navigate to the server directory:

   cd prd-studio/mcp-server

3. Install and start:

   npm install
   node index.js --key=YOUR_API_KEY

Configuration

You can use command line arguments or environment variables:

| Argument | Environment Variable | Description | |----------|----------------------|-------------| | --key | PRD_STUDIO_API_KEY | Your API Key (Required) | | --url | PRD_STUDIO_URL | API Endpoint (Default: https://prd.tdwas.com) | | --project | PRD_STUDIO_PROJECT | Optional: Pin a specific Project ID (Bypasses UI Sync) |

Available Tools

Once connected, your AI agent can use:

• get_active_context: Check which project is currently pinned.
• get_build_roadmap: Plan the implementation sequence.
• get_implementation_prompt: Fetch deep-dive code specs for a feature.
• complete_roadmap_step: Sync progress back to your dashboard.
• ...and more!

Built by PRD Studio