home / mcp / spec workflow mcp server

Spec Workflow MCP Server

Provides a structured, auditable workflow from requirements to tasks for AI-assisted software development.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "kingkongshot-specs-workflow-mcp": {
      "command": "node",
      "args": [
        "/absolute/path/to/specs-mcp/dist/index.js"
      ]
    }
  }
}

You can use the Spec Workflow MCP to guide AI-driven software projects from requirements to design to actionable tasks, with built‑in progress tracking and clear handoffs at every stage. This MCP helps you achieve traceable, organized development that stays aligned with business goals.

How to use

You interact with the Spec Workflow MCP as your project’s central workflow assistant. Describe your needs, and the AI will generate structured documents for requirements, design decisions, and concrete tasks. After each stage, you review and confirm before proceeding, ensuring the project stays on track. Use the persistent progress feature to resume from where you left off in a new session.

How to install

Prerequisites you need before installing the MCP server.

# Prerequisites
node --version  # Node.js should be >= v18.0.0
npm --version     
# Optional: a client that supports MCP integration (Claude, Cursor, Windsurf, etc.)

Install using the Claude Code flow (one common path shown):

claude mcp add spec-workflow-mcp -s user -- npx -y spec-workflow-mcp@latest

Install from source and build the MCP server, then run the generated server file:

git clone https://github.com/kingkongshot/specs-mcp.git
cd specs-mcp
npm install
npm run build

# Start the server after building
# Use the explicit runtime command shown here
node /absolute/path/to/specs-mcp/dist/index.js

Additional sections

Configuration and usage notes from the workflow guide help you manage multi‑module projects, track tasks, and monitor progress. You can place documentation under a dedicated specs folder per module and use a .workflow-confirmations.json file to record status. The workflow supports single-module or multi-module structures, allowing you to organize authentication, payments, notifications, and other domains distinctly.

If you want to systematically guide AI through a project, make sure to enable progress checks, keep language consistent across all docs, and always complete tasks in small, achievable steps (roughly 1–2 hours per task). This keeps the project focused and steadily advancing.

Available tools

specs-workflow

Manages documentation progress and task tracking for the Spec Workflow MCP, enabling sequential progress through requirements, design, and tasks with confirmations between stages.