Dental Lead Qualifier - Python + Claude Code CLI + GLM

Simple Python alternative to the complex n8n workflow.

What it does

Slack Webhook → Extract Lead → SerpAPI Search → Claude Code (GLM) → Slack Reply

Setup

1. Install Python dependencies

pip install -r requirements.txt

2. Configure Claude Code CLI with GLM

Edit ~/.config/claude-code/config.json (or C:\Users\YOURUSER\.claude\config.json on Windows):

{
  "apiProvider": "openai-compatible",
  "apiKey": "6024c744cdc4468b8b337ff4df5038be.mmQhTB8WHG1rSXOp",
  "baseUrl": "https://api.z.ai/api/coding/paas/v4",
  "modelId": "glm-4"
}

3. Set environment variables (optional)

export SLACK_BOT_TOKEN="xoxb-..."
export SERPAPI_KEY="..."
export PORT=5678

Or edit the defaults in lead_qualifier.py.

4. Run the server

python lead_qualifier.py

5. Point Slack webhook to your server

Use Cloudflare Tunnel or ngrok to expose localhost:

cloudflare tunnel --url http://localhost:5678

Then add the webhook URL to your Slack app:

https://your-tunnel-url.cloudflare.net/webhook

Files

• lead_qualifier.py - Main server (reuses all n8n logic)
• requirements.txt - Python dependencies
• README.md - This file

Features

• ✅ Lead extraction (same regex as n8n)
• ✅ SerpAPI dentist verification
• ✅ Claude Code CLI with GLM-4
• ✅ JSON response parsing
• ✅ Slack reply with emoji/status
• ✅ Error handling & logging