Overview
The local Dynatrace MCP server allows AI Assistants to interact with the Dynatrace observability platform, bringing real-time observability data directly into your development workflow. It enables AI agents and MCP clients (e.g., VS Code, Claude, Cursor, Amazon Q, Windsurf, ChatGPT, GitHub Copilot) to execute Dynatrace Query Language (DQL), retrieve logs, events, metrics, spans, and more, and to access problem, vulnerability, and ownership information about monitored entities. It includes built-in cost awareness through Grail query budgets and supports telemetry for usage and errors. The server can run in STDIO mode for integration with popular MCP clients, and in HTTP server mode for stateful sessions and web-based clients. It also provides AI-powered capabilities like NL2DQL and DQL2NL explanations, Slack messaging, and workflow automation via Dynatrace AutomationEngine. Configuration options include environment, proxy settings, and a comprehensive scope set for authentication. It is not officially supported by Dynatrace and is provided for experimentation and local use.
Features
Execute Dynatrace Query Language (DQL)
Run DQL to retrieve logs, events, spans and metrics from Grail and Dynatrace data sources.
Query problem and security data
List and get details about problems and security issues, including vulnerabilities, in monitored services.
Entity insights and ownership
Get more information about monitored entities and their ownership.
Slack and automation integrations
Send Slack messages via Slack Connector and configure notification Workflows via AutomationEngine.
AI-powered assistance (NL2DQL, DQL2NL)
Natural language to DQL, DQL explanations and an AI chat assistant for Dynatrace questions.
Grail budget tracking
Built-in Grail query budget tracking with configurable DT_GRAIL_QUERY_BUDGET_GB and warnings.
HTTP server mode
Run the MCP server as an HTTP service for stateful sessions and web client integration.
Who Is This For?
- Developers:Integrate Dynatrace observability data into AI-assisted workflows and prompts.
- AI assistants:Leverage MCP to query Dynatrace with DQL and receive explanations and insights.
- MCP clients and integrators:Configure and run the MCP server in IDEs and tools like VS Code, Claude, Cursor, Amazon Q, Windsurf, and Gemini.
