​
MCP Integration Examples

Learn how to use the Unhook MCP server effectively with these practical examples and workflows.

​
Common Debugging Workflows

​
Investigating Failed Webhooks

When webhooks start failing, use this workflow to identify and resolve issues:

// 1. Ask Claude to investigate failures
"Show me all failed webhook events from the last hour"

// Claude will use search_events tool:
{
  "name": "search_events",
  "arguments": {
    "status": "failed",
    "limit": 100
  }
}

// 2. Analyze specific failure
"Why did event evt_123 fail?"

// Claude will use analyze_event tool:
{
  "name": "analyze_event", 
  "arguments": {
    "eventId": "evt_123"
  }
}

// 3. Get recommendations
"What should I do to fix these webhook failures?"

​
Performance Monitoring

Track webhook performance and identify bottlenecks:

// Ask: "Generate a performance report for my webhooks"

// The AI will use get_webhook_stats:
{
  "name": "get_webhook_stats",
  "arguments": {
    "timeRange": "24h"
  }
}

// Follow up with:
// "Which webhook has the slowest response time?"
// "Show me the performance trend over the last week"

​
Payload Inspection

Examine webhook payloads for debugging:

// Ask: "Show me the payload for the last GitHub webhook"

// The AI will:
// 1. Read recent events
// 2. Filter for GitHub provider
// 3. Show the payload data

// Or be specific:
// "What data did Stripe send in event evt_xyz?"

​
Advanced Use Cases

​
Failure Pattern Analysis

Identify recurring issues across webhooks:

// Ask: "Find patterns in webhook failures over the last 7 days"

// The AI will:
{
  "name": "get_webhook_stats",
  "arguments": {
    "timeRange": "7d"
  }
}

// Then analyze error distribution and suggest:
// - Common failure causes
// - Time-based patterns
// - Affected endpoints

​
Comparative Analysis

Compare webhook performance:

// Ask: "Compare success rates between my Stripe and PayPal webhooks"

// The AI will gather stats for both:
[
  {
    "name": "get_webhook_stats",
    "arguments": {
      "webhookId": "wh_stripe_prod",
      "timeRange": "24h"
    }
  },
  {
    "name": "get_webhook_stats",
    "arguments": {
      "webhookId": "wh_paypal_prod",
      "timeRange": "24h"
    }
  }
]

​
Event Replay Preparation

Prepare to replay failed events:

// Ask: "Help me replay the failed payment webhooks from today"

// The AI will:
// 1. Search for failed payment events
// 2. Analyze why they failed
// 3. Suggest fixes before replay
// 4. Provide replay instructions

​
Integration Patterns

​
Continuous Monitoring

Set up regular health checks:

// Morning routine:
"Give me a webhook health summary for the last 24 hours"

// The AI provides:
// - Total events processed
// - Success/failure rates
// - Any new error types
// - Performance metrics
// - Recommendations

​
Incident Response

Quick debugging during outages:

// When alerts fire:
"URGENT: Webhooks are failing. What's happening?"

// Rapid diagnosis flow:
// 1. Check recent failures
// 2. Identify error patterns  
// 3. Analyze specific failures
// 4. Provide immediate fixes

​
Development Testing

Validate webhook implementations:

// During development:
"Show me the test webhooks I sent in the last hour"

// Verify payloads:
"Is my endpoint receiving the correct Stripe signature?"

// Debug issues:
"Why is my local endpoint returning 401?"

​
Prompt Engineering Tips

​
Effective Queries

✓ "Show me failed Stripe webhooks from the last hour"
✓ "Why did event evt_123 fail?"
✓ "Compare webhook performance this week vs last week"
✓ "What's causing timeout errors on my GitHub webhook?"

​
Context Building

Provide context for better analysis:

// Instead of: "Why are webhooks failing?"
// Try: "My Stripe webhooks started failing after deploying v2.1. 
//       Show me what changed and how to fix it."

// Instead of: "Show webhook stats"  
// Try: "Generate a performance report for customer-facing webhooks,
//       focusing on payment processing delays"

​
Automation Ideas

​
Daily Reports

// Morning standup prep:
"Generate a daily webhook report with:
- Total events processed
- Failure rate by provider
- New error types since yesterday
- Performance degradation warnings
- Top 3 issues to address"

​
Alert Investigation

// When monitoring alerts trigger:
"Investigate webhook alert: 
- Error rate above 5% for wh_stripe_prod
- Started 15 minutes ago
- Show me what's failing and why"

​
Pre-deployment Checks

// Before deploying:
"Analyze webhook health for the last 2 hours.
Are there any issues I should fix before deploying?"

​
Code Integration Examples

​
Error Handler Improvement

// Ask: "Based on recent failures, how should I improve my webhook error handling?"

// The AI analyzes failures and suggests:
class WebhookHandler {
  async processWebhook(payload: any) {
    try {
      // Validate signature
      if (!this.validateSignature(payload)) {
        // AI notices many 401 errors
        throw new UnauthorizedError('Invalid signature');
      }
      
      // Process with timeout
      // AI suggests timeout based on performance stats
      const result = await Promise.race([
        this.process(payload),
        this.timeout(30000) // 30s based on p99 latency
      ]);
      
      return result;
    } catch (error) {
      // AI suggests specific error handling
      if (error.code === 'ECONNREFUSED') {
        // Retry logic for connection issues
        return this.scheduleRetry(payload);
      }
      throw error;
    }
  }
}

​
Monitoring Integration

// Ask: "Generate monitoring code based on common webhook issues"

// The AI creates:
class WebhookMonitor {
  async checkHealth() {
    // Based on MCP data analysis
    const stats = await this.getWebhookStats('24h');
    
    // Alert on high failure rate
    if (stats.failureRate > 0.05) {
      await this.alert('High failure rate', stats);
    }
    
    // Alert on performance degradation
    if (stats.p95Latency > 5000) {
      await this.alert('Slow webhook processing', stats);
    }
  }
}

​
Best Practices

​
1. Start Broad, Then Narrow

// First: "Show me webhook health overview"
// Then: "Focus on failed Stripe payment webhooks"  
// Finally: "Analyze event evt_123 in detail"

​
2. Use Time Ranges Effectively

// Recent issues: "last hour", "last 24h"
// Trends: "last 7d", "last 30d"
// Comparisons: "this week vs last week"

​
3. Combine Multiple Tools

// Don't just search, also analyze:
"Find failed webhooks AND explain why they failed AND suggest fixes"

​
4. Ask for Actionable Insights

// Instead of: "Show me errors"
// Ask: "What specific changes should I make to reduce webhook failures?"

​
Troubleshooting Examples

​
No Data Appearing

// Diagnose: "Why am I not seeing any webhook data?"
// The AI will check:
// - If webhooks are configured
// - If events are being received
// - If there are permission issues

​
Inconsistent Results

// Ask: "Why do some webhooks work but others fail randomly?"
// The AI analyzes:
// - Success patterns
// - Failure patterns
// - Environmental factors
// - Suggests targeted fixes

​
Next Steps