HireScript MCP Server

MCP (Model Context Protocol) server for generating inclusive, bias-free job descriptions using Claude AI.

Features

• 🎯 AI-Powered Generation - Uses Claude to create professional job descriptions
• ✨ Bias Detection - Analyzes content for gendered, ageist, and ability-biased language
• 📊 Bias Scoring - Returns a 0-100 inclusivity score
• 🔧 MCP Compatible - Works with any MCP-enabled client

Installation

npm install
npm run build

Configuration

Set your Anthropic API key:

export ANTHROPIC_API_KEY=your-api-key

Usage

As MCP Server

Add to your MCP client configuration:

{
  "mcpServers": {
    "hirescript": {
      "command": "node",
      "args": ["/path/to/hirescript-mcp/dist/index.js"],
      "env": {
        "ANTHROPIC_API_KEY": "your-api-key"
      }
    }
  }
}

Tool: generate_job_description

Generate an inclusive job description with bias analysis.

Parameters:

Name	Type	Required	Description
job_title	string	✅	The job title (e.g., "Senior Software Engineer")
company	string	❌	Company name for personalization
requirements	string	❌	Key requirements, one per line
benefits	string	❌	Benefits to highlight
work_mode	string	❌	"remote", "hybrid", or "onsite" (default: remote)

Example Request:

{
  "job_title": "Senior Software Engineer",
  "company": "TechCorp",
  "requirements": "5+ years experience\nTypeScript proficiency\nCloud infrastructure knowledge",
  "benefits": "Competitive salary, health insurance, unlimited PTO",
  "work_mode": "hybrid"
}

Example Response:

{
  "jobDescription": "# Senior Software Engineer\n\n## About the Role\n\nWe're looking for a Senior Software Engineer to join our team...",
  "biasScore": 92,
  "biasWarnings": [
    {
      "original": "young and energetic",
      "suggestion": "motivated and dynamic",
      "reason": "Avoid age-related language that may discourage older candidates"
    }
  ]
}

Development

# Run in development mode
npm run dev

# Build for production
npm run build

# Run tests
npm test

How It Works

1. Takes job details as input
2. Sends a structured prompt to Claude
3. Claude generates an inclusive job description
4. Analyzes the content for potential bias
5. Returns the description with a bias score and warnings

Bias Detection

The server detects and suggests alternatives for:

• Gendered language (he/she → they)
• Ageist terms (young, energetic → motivated, dynamic)
• Ability-biased language (stand for long periods → specific accommodation notes)
• Unnecessary requirements (suggests limiting to truly essential qualifications)
• Exclusionary phrases (rockstar, ninja → high performer, expert)

License

MIT