rahulraonatarajan/prompt-go-mcp

🎯 Prompt Go — Smart AI Workflow for Cursor Teams

Prompt smarter, spend less, work faster.

Transform your team's AI workflow with intelligent routing, real-time cost optimization, and adaptive learning. Prompt Go automatically routes each prompt to the most effective channel while providing actionable insights to reduce costs by 25-40%.

✨ What Makes It Special for Cursor Teams

🧠 Invisible Intelligence

• Zero-friction setup - Works seamlessly in Cursor with one command
• Smart routing - Automatically suggests web/agent/ask/direct based on prompt analysis
• Visual indicators - Subtle UI hints show routing confidence without interrupting flow
• Adaptive learning - Gets smarter from your team's usage patterns

💰 Proactive Cost Management

• Real-time budget tracking with smart alerts before overruns
• Automatic downgrades when approaching limits (soft enforcement)
• Cost predictions based on current usage trends
• Team efficiency scoring - see who's optimizing best

📊 Team Insights Dashboard

• Beautiful analytics at http://localhost:8000/insights/team/your-team
• Savings tracking - see exactly how much smart routing saves
• Weekly recommendations - auto-generated Cursor rules for your team
• Performance benchmarking across team members

🔒 Privacy-First Design

• 100% local processing - no data leaves your infrastructure
• Prompt hashing - content never stored, only patterns analyzed
• Team isolation - each org has separate database and config

🚀 Quick Start for Cursor Teams

Automated Setup (Recommended - 2 minutes)

git clone https://github.com/rahulraonatarajan/prompt-go-mcp.git
cd prompt-go-mcp
python setup-cursor-team.py --org "Your Team Name" --budget 500

That's it! The script will:

• ✅ Install dependencies and setup database
• ⚙️ Configure Cursor MCP settings automatically
• 📋 Create team-specific policy configuration
• 🧪 Test the integration

Manual Setup (if needed)

python -m venv .venv && . .venv/bin/activate
pip install -e .
cp .env.example .env

Add to Cursor settings.json:

{
  "mcpServers": {
    "prompt-go": {
      "command": "python", 
      "args": ["-m", "apps.mcp_server"],
      "cwd": "/absolute/path/to/prompt-go-mcp",
      "env": {
        "DATABASE_URL": "sqlite+aiosqlite:///./promptgo_yourteam.db",
        "PG_ORG": "Your Team Name",
        "PG_ENABLE_LEARNING": "true",
        "PG_BUDGET_MODE": "soft"
      }
    }
  }
}

📊 See It In Action

After setup, try these prompts in Cursor:

• "What's the latest in React 19?" → 🌐 Web suggested
• "Implement OAuth flow for our app" → 🤖 Agent mode
• "Best practices for..." → 💬 Ask for clarification
• "Explain useState hook" → ⚡ Direct answer

Visit your team dashboard: http://localhost:8000/insights/team/your-team-name

🎯 Complete Documentation

📖 - Comprehensive setup and configuration
🔧 - Technical details and troubleshooting
📊 Dashboard: Start with python -m apps.http_dashboard then visit localhost:8000