generated: '2026-08-29'
method: searched
source: >-
  https://www.dynatrace.com/trial/, https://docs.dynatrace.com/docs/dynatrace-api/basics/dynatrace-api-authentication,
  https://docs.dynatrace.com/docs/dynatrace-api/basics/access-limit
provider: Dynatrace
providerId: dynatrace
description: >-
  Dynatrace publishes NO test-mode API. There is no sandbox host, no test-vs-live key prefix pair,
  and no fixture or simulation tooling. What it offers instead is a time-boxed free trial that
  provisions a REAL tenant — so every call an agent makes during evaluation runs against
  production infrastructure and consumes real licence units. This is a meaningful difference from
  a payments-style sandbox and should not be recorded as one.

sandbox_available: false
docs: []
test_mode: false
test_host: null
test_key_prefix: null
live_key_prefix_note: >-
  Dynatrace token prefixes (dt0s01 API token, dt0s02/dt0s03 OAuth clients, dt0s04 chat and
  identity linking, dt0s06 refresh token) distinguish token TYPE, not test versus live. There is
  no test-mode counterpart to any of them.

evaluation_path:
  name: Dynatrace free trial
  url: https://www.dynatrace.com/trial/
  status: 200
  kind: real tenant, time-boxed
  note: >-
    A trial tenant is a production environment with a full API surface. Ingest during a trial is
    real ingest.

cost_awareness:
  note: >-
    Read access to the Dynatrace API is documented as free of charge on a fair-use model, but
    DEFINING and PUSHING new custom metrics is charged per metric per month, and DQL queries
    against Grail consume licence units. An agent exploring this API can therefore incur cost
    without writing any configuration — the Dynatrace MCP README carries the same warning about
    example prompts leading to billable DQL.
  source: https://docs.dynatrace.com/docs/dynatrace-api/basics/access-limit

safe_exploration:
  - Read operations (listEntities, listProblems, listEvents, listMetrics) are the fair-use free path.
  - >-
    The local MCP server shipped a reset_grail_budget tool specifically to cap Grail spend during
    agent exploration; that tool has no remote equivalent, so budget control on the remote MCP
    server is a Dynatrace-side setting rather than a callable guard.

fixtures: []
test_values: []
test_values_note: >-
  No published magic identifiers, test cards, test tokens or simulator values exist for this API.
  Nothing is invented here.

x-evidence:
  - url: https://www.dynatrace.com/trial/
    status: 200
  - url: https://docs.dynatrace.com/docs/dynatrace-api/basics/access-limit
    status: 200
  - url: https://docs.dynatrace.com/docs/dynatrace-api/basics/dynatrace-api-authentication
    status: 200