Best practices for implementing unit tests in this TypeScript monorepo. Use when writing new tests, refactoring existing tests, or fixing failing tests. Covers mocking strategies, test organization, helper functions, and assertion patterns.
Installation
Details
Usage
After installing, this skill will be available to your AI coding assistant.
Verify installation:
skills listSkill Instructions
name: unit-test-implementation description: Best practices for implementing unit tests in this TypeScript monorepo. Use when writing new tests, refactoring existing tests, or fixing failing tests. Covers mocking strategies, test organization, helper functions, and assertion patterns.
Unit Test Implementation Guide
Mocking Dependencies
Use jest-mock-extended for External Dependencies
Use jest-mock-extended for dependencies that are external to the unit under test:
import { mock, mockDeep, mockFn, type MockProxy } from "jest-mock-extended";
let httpClient: MockProxy<HttpClient>;
let database: MockProxy<Database>;
beforeEach(() => {
httpClient = mock<HttpClient>();
database = mockDeep<Database>(); // Use mockDeep when mocking nested properties
});
When to Use Real Instances vs Mocks
Mock external dependencies that are:
- Difficult to set up and not to relevant to the behavior being tested
- External services or APIs
Use real instances for dependencies that are:
- Tightly coupled to the subject (e.g., a config object, a small utility class)
- Pure functions or simple data transformers
- Part of the same module and tested together
When in doubt, ask the user which dependencies should be mocked and which should use real instances. The decision depends on what behavior you're trying to test.
Prefer mockReturnValueOnce Over Complex mockImplementation
// ❌ Bad: Counter-based implementation
let callCount = 0;
mock.fetch.mockImplementation(() => {
callCount++;
return callCount === 1 ? responseA : responseB;
});
// ✅ Good: Clear sequence of return values
mock.fetch
.mockReturnValueOnce(responseA)
.mockReturnValueOnce(responseB)
.mockReturnValue(defaultResponse);
When Mock Behavior Becomes Too Complex, Create a Mock Class
If mocking requires complex state management (tracking multiple calls, conditional responses based on arguments, etc.), create a proper fake implementation instead:
/**
* A fake implementation for testing. Does NOT use jest mocks internally.
* Implements the same interface as the real class.
*/
export class MockCache {
private store = new Map<string, Item>();
// Track calls for assertions
public getCalls: string[] = [];
public setCalls: Array<{ key: string; value: Item }> = [];
async get(key: string): Promise<Item | undefined> {
this.getCalls.push(key);
return this.store.get(key);
}
async set(key: string, value: Item): Promise<void> {
this.setCalls.push({ key, value });
this.store.set(key, value);
}
/** Seed data for testing */
seed(entries: Array<[string, Item]>): this {
entries.forEach(([k, v]) => this.store.set(k, v));
return this;
}
/** Reset for reuse */
reset(): void {
this.store.clear();
this.getCalls = [];
this.setCalls = [];
}
}
Exposing Internals for Testing
Use a Derived Test Class
When you need to access or modify internal state, create a test subclass that exposes protected members:
Base class (production code):
class MyService {
protected config: Config;
constructor(config: Config) {
this.config = config;
}
protected async waitUntilReady(): Promise<void> {
await sleep(1000);
}
async execute(): Promise<Result> {
await this.waitUntilReady();
// ... implementation
}
}
Test class (test file):
class TestMyService extends MyService {
/** Override to skip delays in tests */
protected override async waitUntilReady(): Promise<void> {
return Promise.resolve();
}
/** Expose config for modification */
public updateConfig(partial: Partial<Config>): void {
this.config = { ...this.config, ...partial };
}
/** Expose config for assertions */
public getConfig(): Config {
return this.config;
}
}
Note: This requires the base class to use protected (not private) for members you need to access. Feel free to modify the base class for this when writing its unit tests.
Test Organization
Move Setup Logic to beforeEach
describe("MyFeature", () => {
let subject: TestMyService;
let dependency: MockProxy<Dependency>;
beforeEach(() => {
dependency = mock<Dependency>();
subject = new TestMyService(dependency);
});
// Tests modify subject via update methods, not direct mutation
});
Use Nested beforeEach for Variations
describe("with caching enabled", () => {
beforeEach(() => {
subject.updateConfig({ cacheEnabled: true });
});
it("returns cached results", async () => {
/* ... */
});
});
describe("with caching disabled", () => {
beforeEach(() => {
subject.updateConfig({ cacheEnabled: false });
});
it("always fetches fresh data", async () => {
/* ... */
});
});
Avoid Direct Object Mutation
// ❌ Bad: Direct mutation
config.maxRetries = 5;
// ✅ Good: Use update methods
subject.updateConfig({ maxRetries: 5 });
Helper Functions for Readability
Extract Repeated Setup Patterns
When you find yourself repeating the same 3-4 setup calls across multiple tests, extract them:
// ❌ Bad: Repeated in every test
it("test 1", async () => {
const items = await Promise.all([createItem(1), createItem(2)]);
mockSource.getItems.mockResolvedValue(items);
mockValidator.validate.mockResolvedValue(true);
// ... actual test
});
it("test 2", async () => {
const items = await Promise.all([createItem(1), createItem(2)]);
mockSource.getItems.mockResolvedValue(items);
mockValidator.validate.mockResolvedValue(true);
// ... actual test
});
// ✅ Good: Extracted helper
async function setupValidItems(count: number) {
const items = await Promise.all(times(count, (i) => createItem(i)));
mockSource.getItems.mockResolvedValue(items);
mockValidator.validate.mockResolvedValue(true);
return items;
}
it("test 1", async () => {
const items = await setupValidItems(2);
// ... actual test
});
Extract Repeated Assertion Patterns
When the same assertion sequence appears in multiple tests:
// ❌ Bad: Repeated in every test
it("test 1", async () => {
await subject.publish();
expect(publisher.enqueue).toHaveBeenCalledTimes(1);
expect(publisher.enqueue).toHaveBeenCalledWith(
expect.objectContaining({ type: "proposal" }),
expect.any(Date)
);
expect(publisher.send).toHaveBeenCalled();
});
// ✅ Good: Extracted helper
const expectPublished = () => {
expect(publisher.enqueue).toHaveBeenCalledTimes(1);
expect(publisher.enqueue).toHaveBeenCalledWith(
expect.objectContaining({ type: "proposal" }),
expect.any(Date)
);
expect(publisher.send).toHaveBeenCalled();
};
it("test 1", async () => {
await subject.publish();
expectPublished();
});
Note: Only extract when the pattern repeats across multiple tests. A single complex assertion block doesn't need extraction.
Helper Location
- Suite-specific helpers: Define inside the describe block
- Package-shared helpers: Move to
src/test/utils.ts - Cross-package helpers: Consider if they belong in a shared testing package
Assertions
Be Specific
// ❌ Too generic - doesn't verify actual behavior
expect(mock.save).toHaveBeenCalledWith(expect.anything());
// ✅ Specific - verifies the important parts
expect(mock.save).toHaveBeenCalledWith(
expect.objectContaining({
id: expectedId,
status: "active",
})
);
Use Mock Class Properties for Assertions
When using custom mock classes, assert on tracked properties:
expect(mockCache.getCalls).toEqual(["key1", "key2"]);
expect(mockCache.setCalls).toHaveLength(1);
expect(mockCache.setCalls[0]).toEqual({ key: "key1", value: expectedValue });
Reusable Test Suites
For testing multiple implementations of the same interface (e.g., in-memory vs persistent storage), create a shared test suite:
// cache_test_suite.ts
export function describeCacheImplementation(
name: string,
createCache: () => Cache
) {
describe(name, () => {
let cache: Cache;
beforeEach(() => {
cache = createCache();
});
it("stores and retrieves values", async () => {
await cache.set("key", "value");
expect(await cache.get("key")).toBe("value");
});
it("returns undefined for missing keys", async () => {
expect(await cache.get("missing")).toBeUndefined();
});
});
}
// memory_cache.test.ts
describeCacheImplementation("MemoryCache", () => new MemoryCache());
// redis_cache.test.ts
describeCacheImplementation("RedisCache", () => new RedisCache(mockClient));
Testing Async Operations
Use Promise.allSettled for Multiple Outcomes
it("handles mixed success and failure", async () => {
const results = await Promise.allSettled([
subject.processValid(),
subject.processInvalid(),
]);
expect(results).toEqual([
{ status: "fulfilled", value: expectedValue },
{
status: "rejected",
reason: expect.objectContaining({ message: "Invalid" }),
},
]);
});
Control Time in Tests
beforeEach(() => {
jest.useFakeTimers();
});
afterEach(() => {
jest.useRealTimers();
});
it("retries after delay", async () => {
mock.fetch.mockRejectedValueOnce(new Error("fail")).mockResolvedValue(data);
const promise = subject.fetchWithRetry();
await jest.advanceTimersByTimeAsync(1000);
await expect(promise).resolves.toEqual(data);
});
Running Tests
cd <package-name>
yarn build # Always compile first
yarn test my-class.test.ts # Run specific test file
yarn test my-class.test.ts -t 'test name' # Run specific test
env LOG_LEVEL=verbose yarn test my-class.test.ts # With debug logging
More by AztecProtocol
View allSpawn an independent Claude instance in a git worktree to work on a task in parallel. Use when the user wants to delegate a task to run independently while continuing the current conversation.
Investigate CI failures by following log links from ci.aztec-labs.com. Given a GitHub Actions URL, PR number, or direct CI log URL, fetch logs and recursively follow nested log links to trace the root cause of build or test failures.
Updates changelog documentation for contract developers and node operators by analyzing branch changes relative to 'next'. Use when preparing a PR, updating migration notes, documenting breaking changes, or when asked to update changelog/release notes.
Guidelines for writing module READMEs that explain how a module works to developers who need to use it or understand its internals. Use when documenting a module, package, or subsystem.
