LoginSign Up
← Back to all sparks
M

Microcks

INFRA · APIS
Velocity0.0

API mocking and contract testing platform

microcks.io ↗

Microcks is quietly wiring LLM example generation into an API mocking tool built for Kubernetes.

api-mockingcontract-testingkubernetesnative-imagellm-generationasync-api
◆Current state
Three release candidates are visible — 1.14.0-rc1 and rc2 in March 2026, then 1.15.0-rc1 in June. The 1.14 line added Sync-to-Async triggers with trigger info propagated to the async minion, pushed the Helm chart as an OCI artifact, fixed API health checks under bad connection configuration, and extended OpenAI-based example generation to the GPT-5 family. 1.15.0-rc1 is largely platform upkeep: Angular upgraded to v21, netty explicitly embedded in the native image, GraphQL extensions, paginated topic ARN lookup, and a corrected OpenAPI overlay export shape.
◆Where it's heading
Two threads run in parallel. The deployment story keeps getting tightened for Kubernetes — OCI-packaged Helm charts, native-image reflection registration, an Uber distribution — which is where a mock server has to live to be useful in CI. The more interesting thread is example generation: authoring realistic mock data is the actual work of using a mocking tool, and Microcks is putting a model behind it rather than asking users to write it. That work is still surfacing as commit lines rather than headline features.
◆Prediction
The 1.15 line should reach a final release given the pre-release pattern, and the LLM example generation is the thread most likely to move from plumbing to a named capability if it keeps expanding.

◆Recent moves

  1. 1mo ago

    1.15.0-rc1: Angular 21, native-image netty fix, GraphQL extensions

    Platform maintenance across the stack — the UI moves to Angular v21, netty is embedded explicitly so the native image stops throwing, and a default metrics URL that was initialising a registry it should not have is removed. The OpenAPI overlay export shape correction is the one change that alters output users consume.

    View source ↗
  2. 4mo ago

    1.14.0-rc2 extends OpenAI example generation to GPT-5 models

    Mostly a housekeeping candidate, but it carries the model-family expansion for OpenAI-backed example generation and adds explain tracing to operation service errors. The generation feature broadening rather than deepening puts it in line with the incremental pattern across this cycle.

    View source ↗
  3. 4mo ago

    1.14.0-rc1 adds Sync-to-Async triggers and OCI-packaged Helm charts

    The substantive candidate of the 1.14 line: Sync-to-Async triggers with trigger info propagated to the async minion let a synchronous call drive event-based mocks, closing a gap between the two protocol worlds Microcks covers. Shipping the Helm chart as an OCI artifact and registering reflection classes for native builds continue the Kubernetes deployment work.

    View source ↗