import requests
import json
from datetime import datetime, timedelta

# Configuration
ZENDESK_DOMAIN = "your-company.zendesk.com"
ZENDESK_API_KEY = "your_api_key"
SLACK_WEBHOOK = "https://hooks.slack.com/services/YOUR/WEBHOOK/URL"

def fetch_zendesk_tickets():
    """Get all tickets updated in the last 24 hours"""
    yesterday = (datetime.now() - timedelta(days=1)).isoformat()
    url = f"https://{ZENDESK_DOMAIN}/api/v2/tickets.json"

    params = {
        "query": f'updated>={yesterday}',
        "per_page": 100
    }

    headers = {
        "Authorization": f"Basic {ZENDESK_API_KEY}",
        "Accept": "application/json"
    }

    response = requests.get(url, params=params, headers=headers)
    return response.json()['tickets']

def fetch_slack_feedback():
    """Get messages from #feedback channel"""
    # Requires Slack SDK: pip install slack-sdk
    from slack_sdk import WebClient

    client = WebClient(token="xoxb-your-slack-token")
    yesterday = int((datetime.now() - timedelta(days=1)).timestamp())

    messages = client.conversations_history(
        channel="C1234567890",  # #feedback channel ID
        oldest=yesterday
    )
    return messages['messages']

def compile_feedback():
    """Combine all sources"""
    zendesk_tickets = fetch_zendesk_tickets()
    slack_messages = fetch_slack_feedback()

    feedback_list = []

    for ticket in zendesk_tickets:
        feedback_list.append({
            "source": "zendesk",
            "id": ticket['id'],
            "content": f"{ticket['subject']}: {ticket['description']}",
            "timestamp": ticket['updated_at']
        })

    for msg in slack_messages:
        feedback_list.append({
            "source": "slack",
            "id": msg['ts'],
            "content": msg['text'],
            "timestamp": msg['ts']
        })

    return feedback_list

if __name__ == "__main__":
    feedback = compile_feedback()
    print(json.dumps(feedback, indent=2))
    # Next step: send to AI layer

You are a product insights analyst. You will receive customer feedback from multiple channels (support tickets, reviews, Slack messages). Your job is to extract structured insights.

For each piece of feedback:
1. **Category**: What area of the product does this relate to? (e.g., Pricing, Onboarding, Performance, UX, Feature Request, Bug Report, Competitor Mention)
2. **Sentiment**: Positive, Neutral, Negative, or Mixed
3. **Urgency**: Critical (blocking customers), High (common complaint), Medium (isolated issue), Low (nice-to-have feedback)
4. **Customer Segment**: Enterprise, Mid-Market, Startup, Agency, or Unknown
5. **Exact Quote**: The most important 1-2 sentences from the feedback
6. **Action Item**: What should the product team do about this? (if applicable)

Process this batch of feedback and return valid JSON:

[FEEDBACK BATCH HERE]

Return only valid JSON in this format:
{
  "insights": [
    {
      "feedback_id": "string",
      "source": "zendesk|slack|review|nps",
      "category": "string",
      "sentiment": "positive|neutral|negative|mixed",
      "urgency": "critical|high|medium|low",
      "customer_segment": "string",
      "quote": "string",
      "action_item": "string or null"
    }
  ],
  "summary": {
    "total_feedback_items": number,
    "top_3_themes": ["string", "string", "string"],
    "critical_issues": [string],
    "competitor_mentions": [string],
    "feature_requests": [string]
  }
}

import anthropic

def process_feedback_with_ai(feedback_list):
    """Send feedback batch to Claude for analysis"""
    client = anthropic.Anthropic(api_key="your_api_key")

    # Format feedback for Claude
    feedback_text = "\n\n".join([
        f"Source: {f['source']}\nID: {f['id']}\nContent: {f['content']}"
        for f in feedback_list
    ])

    message = client.messages.create(
        model="claude-3-5-sonnet-20241022",
        max_tokens=2048,
        messages=[
            {
                "role": "user",
                "content": f"""You are a product insights analyst. You will receive customer feedback from multiple channels (support tickets, reviews, Slack messages). Your job is to extract structured insights.

For each piece of feedback:
1. **Category**: What area of the product does this relate to? (e.g., Pricing, Onboarding, Performance, UX, Feature Request, Bug Report, Competitor Mention)
2. **Sentiment**: Positive, Neutral, Negative, or Mixed
3. **Urgency**: Critical (blocking customers), High (common complaint), Medium (isolated issue), Low (nice-to-have feedback)
4. **Customer Segment**: Enterprise, Mid-Market, Startup, Agency, or Unknown
5. **Exact Quote**: The most important 1-2 sentences from the feedback
6. **Action Item**: What should the product team do about this? (if applicable)

Process this batch of feedback and return valid JSON:

{feedback_text}

Return only valid JSON with no markdown formatting:
{{
  "insights": [
    {{
      "feedback_id": "string",
      "source": "zendesk|slack|review|nps",
      "category": "string",
      "sentiment": "positive|neutral|negative|mixed",
      "urgency": "critical|high|medium|low",
      "customer_segment": "string",
      "quote": "string",
      "action_item": "string or null"
    }}
  ],
  "summary": {{
    "total_feedback_items": number,
    "top_3_themes": ["string", "string", "string"],
    "critical_issues": ["string"],
    "competitor_mentions": ["string"],
    "feature_requests": ["string"]
  }}
}}"""
            }
        ]
    )

    # Parse response
    import json
    return json.loads(message.content[0].text)

def send_slack_digest(insights):
    """Post daily digest to Slack"""
    critical_issues = [i['quote'] for i in insights['insights']
                       if i['urgency'] == 'critical']
    feature_requests = insights['summary']['feature_requests']
    themes = insights['summary']['top_3_themes']

    message = f"""📊 *Daily Feedback Digest*

🚨 *Critical Issues ({len(critical_issues)})*
{chr(10).join([f"• {issue}" for issue in critical_issues[:3]])}

✨ *Top Feature Requests*
{chr(10).join([f"• {req}" for req in feature_requests[:3]])}

📈 *Emerging Themes*
{chr(10).join([f"• {theme}" for theme in themes])}

See full report: https://your-dashboard.com/feedback"""

    requests.post(SLACK_WEBHOOK, json={"text": message})

import smtplib
from email.mime.text import MIMEText

def send_email_digest(insights, recipient_email):
    """Send digest via email"""
    subject = "Daily Feedback Digest"
    body = f"""
    Critical Issues: {len([i for i in insights['insights'] if i['urgency'] == 'critical'])}
    Top Themes: {', '.join(insights['summary']['top_3_themes'])}

    Full details attached or visit your dashboard.
    """

    msg = MIMEText(body)
    msg['Subject'] = subject
    msg['From'] = "feedback@yourcompany.com"
    msg['To'] = recipient_email

    # Send via your mail service

<!DOCTYPE html>
<html>
<head>
    <title>Feedback Dashboard</title>
</head>
<body>
    <h1>Today's Feedback Summary</h1>
    <div id="digest"></div>

    <script>
        fetch('/api/feedback-digest')
            .then(r => r.json())
            .then(data => {
                document.getElementById('digest').innerHTML = `
                    <p>Critical Issues: ${data.summary.critical_issues.length}</p>
                    <p>Top Themes: ${data.summary.top_3_themes.join(', ')}</p>
                `;
            });
    </script>
</body>
</html>