MCP规约分析工具

Prerequisites

• Node.js 18+ • pnpm (recommended)

� Installation & Setup

git clone https://github.com/orangebread/speclinter-mcp.git
cd speclinter-mcp
pnpm install && pnpm build

🚀 Setup Guide

Installation

pnpm install && pnpm build

✅ Verify Setup

```bash

Clone and build

git clone https://github.com/orangebread/speclinter-mcp.git cd speclinter-mcp docker build -t speclinter/mcp .

🚀 Getting Started

"Initialize SpecLinter in my project"
"Analyze my codebase to understand patterns and tech stack"

Docker development

docker build -t speclinter/mcp:latest . docker run -it --rm -v $(pwd):/workspace speclinter/mcp ```

🎯 Quick Start (5 Minutes)

Example: F → A Transformation

❌ Grade F Spec:

"Add social features to the app"

✅ Grade A Spec:

As a user, I want to send direct messages to other users so that I can communicate privately.

Acceptance Criteria:
- Users can search for other users by username
- Messages are delivered in real-time using WebSocket
- Message history is preserved and searchable
- Users can block/unblock other users
- All messages are encrypted end-to-end

Error Handling:
- Show "User not found" for invalid usernames
- Display "Message failed to send" with retry option
- Handle offline users with message queuing

💬 How to Use SpecLinter

Just talk to your AI assistant naturally! Here are the most common commands:

🏗️ Real-World Examples

Example 1: E-commerce Product Catalog

Input Spec:

Create a product catalog page with filtering and search

SpecLinter Output: - Grade: C (needs more detail) - Generated Tasks: 6 structured tasks - Improvements: Add specific filter types, search behavior, pagination

Example 2: User Authentication System

Input Spec:

As a user, I want to log in with email and password so I can access my account.

Acceptance Criteria:
- Email validation with proper error messages
- Password must be 8+ characters
- Failed attempts are rate-limited (3 attempts/5 minutes)
- Successful login redirects to dashboard
- Remember me option for 30 days

SpecLinter Output: - Grade: A (excellent specification) - Generated Tasks: 12 detailed tasks including security, validation, and UX - Gherkin Scenarios: Auto-generated for each task

🔌 MCP Configuration

Add this to your AI IDE's MCP configuration:

{
  "mcpServers": {
    "speclinter": {
      "command": "node",
      "args": ["/absolute/path/to/dist/cli.js", "serve"]
    }
  }
}

🔧 Development Reference

These commands are retained for historical reference for anyone exploring or forking the project.

```bash git clone https://github.com/orangebread/speclinter-mcp.git cd speclinter-mcp

🎯 Core Workflows

MCP Integration

Add this configuration:

{
  "mcpServers": {
    "speclinter": {
      "command": "node",
      "args": ["/absolute/path/to/dist/cli.js", "serve"]
    }
  }
}

📖 GitMCP Integration

For documentation access and examples, you can also add GitMCP integration:

{
  "mcpServers": {
    "speclinter-docs": {
      "command": "npx",
      "args": ["mcp-remote", "https://gitmcp.io/orangebread/speclinter-mcp"]
    }
  }
}

🐳 Docker Alternative

Quick Docker Setup: ```bash