rj-s-org/mockingbird

Mock any API from its contract with Mockingbird - REST/OpenAPI, GraphQL, gRPC, SOAP, SSE, WebSocket, Kafka/AMQP/MQTT - with schema-valid, LLM-generated responses and per-session state. Use when the user wants a mock server, a fake or stub backend, an API sandbox, test doubles for an HTTP/gRPC/GraphQL/SOAP dependency, or to work against an API that does not exist yet. Start here; it routes to the setup, mock-api, deploy and testing skills.

Self-host Mockingbird beyond a laptop - a background server, Docker, the Docker Compose stack (Postgres, OIDC), Kubernetes with the Helm chart (all-in-one or split control/data planes), authentication and API keys, and exposing the MCP server over HTTP for ChatGPT and other remote agents. Use when the user wants a shared, persistent, team or cloud mock endpoint.

Create and drive Mockingbird mocks - import an OpenAPI/Swagger, GraphQL, protobuf/gRPC, WSDL/SOAP or AsyncAPI contract, or turn API docs, a HAR recording or a Postman collection into one; call the mock; add scenarios (forced errors, latency, fields), per-session state and fixtures, chaos, and event channels. Use when the user asks to mock, fake or stub an API or a specific endpoint behaviour.

Use Mockingbird mocks in automated tests and CI - deterministic offline mode, seeded fixtures, isolated sessions per test, forced errors and latency, scenarios, asserting on the request log, and GitHub Actions setup. Use when the user writes integration or end-to-end tests against an external API, or wants a mock dependency in a pipeline.