I’m exploring the use of the Dynatrace MCP server to query observability data using AI and DQL translation features.

Currently, MCP seems to support connecting to one Dynatrace environment per running process, using the DT_ENVIRONMENT and DT_PLATFORM_TOKEN environment variables.

I wanted to confirm:

• Is it possible to configure multiple Dynatrace environments in the same MCP instance (e.g., prod, qa, dev)?

• Or does each environment still require a separate MCP process?

• If single-process multi-environment support is not available, are there any plans or design considerations for enabling it (for example, environment selection per query or via context switching)?

Use case:
I manage multiple Dynatrace tenants (Prod, QA, and Dev) and would like a unified MCP layer to handle all environments, instead of running separate MCP instances for each.