Skip to main content

Testing Guide

Merlon has tests at every layer: Go, Rust, TypeScript, and Proto. This document covers how to run each and the overall testing strategy.

Running tests per component​

Go (API)​

cd api && go test ./...

With coverage:

cd api && go test -cover ./...

Rust (Engine)​

cd engine && cargo test

UI (TypeScript / React)​

cd ui && npm run test

Proto (Protocol Buffers)​

cd proto && buf lint

Detecting breaking changes:

cd proto && buf breaking --against '.git#branch=main'

Running everything​

make test

make test runs the Go tests, Rust tests, UI tests, and buf lint together. CI uses the same target.

Testing strategy (TDD)​

Feature work and bug fixes follow TDD (test-driven development).

  1. Write the test — write a test that expresses the expected behavior first (confirm it fails).
  2. Implement — write the minimal code needed to make the test pass.
  3. Refactor — clean up while keeping the test green.

Emphasis by layer​

  • Engine (Rust) — CDD scoring, TM evaluation, and screening are the core business logic. To uphold reproducibility of decision rationale (Auditability First), tests pin the determinism of output for identical input.
  • API (Go) — CRUD and orchestration. Tests verify service-boundary contracts and that failures fall toward alerting rather than silence (Fail-Alert).
  • Proto — Contract Stability (12+ months of backward compatibility) is mechanically enforced with buf breaking.
  • UI — component-level tests plus user-flow tests for the investigation workflow.

Follow existing patterns​

New tests should follow the existing test framework, naming conventions, and patterns in each directory. Go uses the standard testing package, Rust uses #[cfg(test)] modules, and the UI follows its existing test-runner configuration.