← ClaudeAtlas

grpc-mocklisted

Use when gRPC client code needs tests against a controllable fake server, or a gRPC interceptor needs tests: Go bufconn and mockgen, Python pytest-grpc, JVM in-process servers, Node fake servers, with success, error and timeout responses. Streaming sequences are grpc-streaming-test-author.
testland/qa-skills · ★ 0 · Testing & QA · score 72
Install: claude install-skill testland/qa-skills
# grpc-mock ## Overview Mocking a gRPC server lets client-side tests exercise success paths, every `grpc.StatusCode` (per the status-code catalog in `grpc-streaming-test-author`, references/status-codes.md), timeouts, and streaming sequences without a real backend. The in-process harness is also the substrate for interceptor tests - see [references/interceptors.md](references/interceptors.md). Three approaches dominate, picked by language: | Approach | Mechanism | |---|---| | **In-process gRPC server** | A real `grpc.Server` listens on an in-memory transport (`bufconn` in Go, `InProcessServerBuilder` in JVM). Tests exercise the full client stack. | | **Interface mock** | `mockgen` / `gomock` (Go) / Mockito (JVM) / `unittest.mock` (Python) replace the generated client stub with a programmable mock. Faster but skips marshalling. | | **Standalone mock server** | Run a tool like `grpcmock` / `dishwasher` as a subprocess. Cross-language client testing. | ## When to use - Client-side tests need controllable gRPC responses (success + error matrix). - Streaming-RPC tests need a server that sends a deterministic sequence. - Mid-test failure injection (deadline / cancellation / status code). - Tests must not depend on a running backend (offline CI). - Cross-service contract testing: want the client to see what the contract says, not what the current server happens to do. ## Authoring ### Go: bufconn + in-process server Per [pkg.go.dev/google.golang.org/grpc/test/bufcon