generated: '2026-08-27'
method: searched
source: >-
  https://docs.confluent.io/cloud/current/client-apps/testing.html,
  https://docs.confluent.io/platform/current/get-started/platform-quickstart.html,
  https://www.confluent.io/confluent-cloud/tryfree/,
  https://docs.confluent.io/kafka-connectors/datagen/current/overview.html
description: >-
  Confluent has no test-mode/live-mode key split and no hosted sandbox tenant — the pattern
  familiar from payments APIs does not apply here. What Confluent ships instead is a full local
  substitute for the platform (Docker, embedded clusters, in-process mocks) plus a free tier on
  the real service. An agent evaluating Confluent should expect to run a real cluster, not a
  simulator behind a test key.
test_mode:
  separate_keys: false
  key_prefixes: null
  note: >-
    Confluent API keys are Cloud, Global or Resource-scoped — a scope axis, not a test/live
    axis. There is no test-mode credential; every key acts on real resources. The blast-radius
    control is RBAC scope and a throwaway environment, not a sandbox flag.
free_tier:
  url: https://www.confluent.io/confluent-cloud/tryfree/
  http_status: 200
  description: >-
    Free signup with trial credit, plus the first eCKU free on Basic clusters on an ongoing
    basis. See plans/confluent-plans-pricing.yml.
local_environments:
  - name: Confluent Platform quick start (Docker)
    docs: https://docs.confluent.io/platform/current/get-started/platform-quickstart.html
    description: >-
      Full local Confluent Platform with Docker Compose — Kafka, Connect, Schema Registry and
      Control Center. Requires Docker 1.11+ and at least 6 GB allocated to Docker on macOS.
    repo: https://github.com/confluentinc/cp-all-in-one
  - name: Confluent Platform quick start (ZIP/TAR)
    docs: https://docs.confluent.io/platform/current/installation/installing_cp/zip-tar.html
  - name: confluent local
    docs: https://docs.confluent.io/confluent-cli/current/command-reference/local/
    description: CLI command group for running a local Confluent Platform for development.
unit_test_tooling:
  - {name: TopologyTestDriver, scope: Kafka Streams, description: 'Test Kafka Streams topology logic without a running cluster; single-message stepping.', docs: 'https://docs.confluent.io/platform/current/streams/developer-guide/test-streams.html'}
  - {name: MockProcessorContext, scope: Kafka Streams, description: Mocked context that captures data forwarded by a custom Processor.}
  - {name: MockProducer, scope: JVM client, description: 'Implements the KafkaProducer interface, mocking all I/O.'}
  - {name: MockConsumer, scope: JVM client, description: 'Implements the KafkaConsumer interface, mocking all I/O.'}
  - {name: MockSchemaRegistryClient, scope: Schema Registry, description: 'Simulate a Schema Registry instance in unit tests; serialize and deserialize without a live registry.', repo: 'https://github.com/confluentinc/schema-registry'}
  - {name: ksql-test-runner, scope: ksqlDB, description: 'Command-line test runner that checks ksqlDB statements against given inputs with no running Kafka or ksqlDB cluster.', caveat: 'Confluent states it can change and has no backward compatibility guarantees.'}
  - {name: rdkafka_mock, scope: non-JVM (librdkafka), description: 'Minimal implementation of the Kafka protocol broker APIs with no other dependencies.'}
integration_test_tooling:
  - {name: Testcontainers for Kafka, description: 'Confluent''s recommended integration-test path — brokers in Docker with isolated JVM, memory and CPU.', url: 'https://www.testcontainers.org/modules/kafka/'}
  - {name: EmbeddedKafkaCluster, description: 'Asynchronous multi-threaded cluster in the same JVM as the tests, from kafka-streams-test-utils.', caveat: 'Confluent states it is not a public interface and has no backward compatibility guarantees.'}
  - {name: EmbeddedSingleNodeKafkaCluster, description: 'As EmbeddedKafkaCluster but including Schema Registry.', repo: 'https://github.com/confluentinc/kafka-streams-examples'}
  - {name: Trivup, description: 'Light-weight community-supported framework for bringing up a simple Kafka cluster (non-JVM clients).'}
fixture_and_trigger_tooling:
  - name: Datagen Source Connector
    docs: https://docs.confluent.io/kafka-connectors/datagen/current/overview.html
    description: >-
      Confluent's first-party mock-data generator. Produces synthetic records to topics using
      predefined or custom Avro schemas — the closest thing Confluent has to seedable fixtures.
  - name: Advanced Message Search
    description: >-
      Console message browser filter for inspecting retained history by field value (Open
      Preview since 2026-08-12). Useful for verifying test writes without a consumer app.
schema_compatibility_testing:
  - {name: Schema Registry Maven plugin, docs: 'https://docs.confluent.io/cloud/current/sr/develop/maven-plugin.html', description: 'Checks a new schema for compatibility against previously registered schemas as a build step.'}
  - {name: testCompatibilityBySubjectName, kind: api-operation, description: 'Schema Registry operation that tests a candidate schema against a subject''s compatibility rules before registering it — a genuine dry-run for the one write surface that has one.'}
test_values:
  published: false
  note: >-
    Confluent publishes no magic test identifiers, test cards or reserved sentinel values.
    Nothing is transcribed here rather than invented.
agent_guidance: >-
  There is no safe rehearsal mode on the live Confluent Cloud APIs. The only true dry-run in the
  whole surface is Schema Registry's compatibility test. For anything else, an agent should
  operate against a dedicated throwaway environment created with createOrgV2Environment and
  torn down afterwards, with a resource-scoped API key — that environment IS the sandbox.