Stop Context Switching Between Your IDE and Production Monitoring

Your production systems are telling you something important, but you're buried in code. Instead of alt-tabbing to Dynatrace dashboards every time something breaks, what if your AI assistant could pull that observability data directly into your development context?

The Dynatrace MCP server bridges this gap by connecting your IDE's AI tools (Claude, Copilot, Cursor) directly to your production telemetry. No more screenshots of error graphs or copying stack traces between windows.

The Problem: Monitoring Lives in a Different World

You're debugging a performance issue. Your current workflow probably looks like this:

1. Notice something's wrong in your code
2. Open Dynatrace in another tab
3. Hunt through dashboards for relevant data
4. Copy/paste error details back to your IDE
5. Try to correlate the production data with your local code
6. Repeat this dance for every investigation

This friction kills productivity and breaks your flow state.

What This Changes

Instead of that workflow, you can now ask your AI assistant directly:

"Our checkout process is slow - show me the full request trace and identify the bottleneck"

Your AI pulls real production traces, correlates them with your current codebase, and gives you actionable insights without leaving your editor.

"Check if this dependency vulnerability I'm seeing locally is actually exploited in production"

Get immediate answers about whether that scary CVE warning actually matters for your running services.

Real Production Context in Your Development Flow

This MCP server gives your AI assistant access to:

• Live problems and incidents from your services
• Security vulnerabilities and their production impact
• DQL query execution for custom telemetry analysis
• Log correlation with your current code changes
• Automated alerting setup via Slack workflows
• Service ownership and team responsibility mapping

All accessible through natural language queries in your IDE.

Concrete Use Cases That Save Hours

Memory leak investigation:

"There's a memory issue on our payment service host. 
Get the problem details and pull related logs to show 
what's causing the spike and which code is responsible."

Deployment troubleshooting:

"Our Kubernetes deployments are failing intermittently. 
Fetch recent events from production-cluster and correlate 
with our deployment configs in this repo."

Security response:

"I'm seeing a vulnerability warning for this dependency. 
Check production for any active exploits and set up 
alerts for the security team if we're actually at risk."

Performance debugging:

"These 503 errors started after yesterday's release. 
Pull error correlations with service health and show me 
which recent commits might be causing circuit breaker trips."

How It Fits Your Stack

Add this to VS Code, Claude Desktop, Cursor, or Amazon Q Developer with a simple config:

{
  "servers": {
    "dynatrace": {
      "command": "npx",
      "args": ["-y", "@dynatrace-oss/dynatrace-mcp-server@latest"],
      "env": {
        "DT_ENVIRONMENT": "https://your-tenant.apps.dynatrace.com",
        "OAUTH_CLIENT_ID": "your-client-id",
        "OAUTH_CLIENT_SECRET": "your-client-secret"
      }
    }
  }
}

Your AI assistant immediately gains access to your production observability data. No additional authentication flows or complex integrations.

Beyond Reactive Debugging

This isn't just about fixing problems faster. You can:

• Validate performance assumptions before code review
• Understand user impact of the feature you're building
• Set up proactive monitoring for new services during development
• Correlate code changes with production metrics in real-time
• Get security context for dependency decisions

Production-Ready Setup

Built by Dynatrace's open source team with proper OAuth integration, comprehensive API coverage, and support for both SaaS and managed deployments. The server handles authentication, rate limiting, and error handling so your AI interactions stay smooth.

You'll need a Dynatrace OAuth client with the right scopes (the README covers the full list), but setup takes minutes, not hours.

Why This Matters Now

Modern development already happens with AI assistance. But that AI is blind to your production reality. This MCP server gives your AI assistant the observability superpowers that turn generic code suggestions into production-aware insights.

Your debugging sessions become conversations with someone who actually knows how your code behaves in the wild.

Ready to stop alt-tabbing to monitoring dashboards? Install @dynatrace-oss/dynatrace-mcp-server and start asking your AI about production directly from your IDE.