feat: Unit test POC for API and Expo layers with coverage infrastructure by Copilot · Pull Request #1914 · PackRat-AI/PackRat

Copilot · 2026-03-09T11:37:33Z

Original prompt

This section details on the original issue you should resolve

<issue_title>Feature: Comprehensive Unit Test Coverage for API and Expo Layers</issue_title>
<issue_description>## Feature Request: Comprehensive Unit Test Coverage for API and Expo Layers

Problem Statement
PackRat currently has significant gaps in unit test coverage across both the API (backend) and Expo (mobile app) layers. Many core business logic paths are untested, leading to regressions and reduced confidence in deployments.

Current State

API layer: ~30% coverage (many service functions untested)
Expo layer: ~25% coverage (component logic, hooks, utilities largely untested)
Integration tests exist but fail silently (see [QA] 18 integration test files silently register 0 tests instead of skipping #1865)

Proposed Solution
Implement comprehensive unit test suite achieving 80%+ coverage for critical paths in both layers.

API Layer Tests

Target Coverage: 80%+

Core Modules to Cover:

Services - Business logic, data transformation, external API calls
- TripService (create, update, delete, get trips)
- PackService (pack CRUD, item management)
- UserService (auth, profile, preferences)
- CatalogService (item search, filtering)
- AIService (LLM interactions, prompt handling)
Controllers - Request/response handling, validation
- All route handlers with mocked services
- Error handling paths
- Input validation (Zod schemas)
Middleware - Auth, logging, rate limiting
- JWT validation
- Request logging
- Error middleware
Utilities - Helpers, formatters, validators
- Date/time utilities
- String/format helpers
- Calculation utilities (weight, distance)
Database/Models - Schema validation, relationships
- Prisma model validation
- Relationship queries

Testing Stack:

Framework: Vitest (already configured)
Mocking: vi.mock() for external deps
DB: @prisma/client mock or test database
Coverage: c8 or @vitest/coverage-v8

Sample Test Pattern:

// packages/api/src/services/__tests__/tripService.test.ts
import { describe, it, expect, vi } from 'vitest';
import { TripService } from '../tripService';
import { prisma } from '../../db';

vi.mock('../../db', () => ({
  prisma: {
    trip: {
      create: vi.fn(),
      findMany: vi.fn(),
      update: vi.fn(),
      delete: vi.fn(),
    },
  },
}));

describe('TripService', () => {
  describe('createTrip', () => {
    it('should create a trip with valid data', async () => {
      const mockTrip = { id: '1', name: 'Test Trip', ... };
      prisma.trip.create.mockResolvedValue(mockTrip);
      
      const result = await TripService.createTrip({ ... });
      
      expect(result).toEqual(mockTrip);
      expect(prisma.trip.create).toHaveBeenCalledWith({ ... });
    });
    
    it('should throw error for invalid dates', async () => {
      await expect(
        TripService.createTrip({ startDate: 'invalid' })
      ).rejects.toThrow('Invalid date format');
    });
  });
});

Expo Layer Tests

Target Coverage: 75%+

Core Areas to Cover:

Hooks - Custom React hooks
- useTrips (fetching, caching, mutations)
- usePacks (CRUD operations)
- useAuth (session, login, logout)
- useOfflineSync (background sync logic)
- useLocalStorage (persistence)
Components - Reusable UI components
- Button, Input, Card variants
- TripCard, PackCard (display logic)
- Form components (validation, submission)
- Modal/Sheet components
Screens - Page-level components
- TripListScreen (filtering, sorting)
- PackDetailScreen (item management)
- ProfileScreen (form handling)
- Test with mocked navigation and context
Utils/Helpers - Pure functions
- Date formatting
- Weight calculations
- URL builders
- Validation functions
State Management - Store logic
- Zustand store actions
- State selectors
- Persistence logic

Testing Stack:

Framework: Jest (Expo's default) or Vitest
React Testing Library: @testing-library/react-native
Mocking: jest.mock for native modules
Navigation: Mocked React Navigation

Sample Test Pattern:

// apps/expo/hooks/__tests__/useTrips.test.ts
import { renderHook, waitFor } from '@testing-library/react-native';
import { useTrips } from '../useTrips';
import { QueryClient, QueryClientProvider } from '@tanstack/react-query';

const queryClient = new QueryClient();

const wrapper = ({ children }) => (
  <QueryClientProvider client={queryClient}>{children}</QueryClientProvider>
);

describe('useTrips', () => {
  it('should fetch trips successfully', async () => {
    const { result } = renderHook(() => useTrips(), { wrapper });
    
    await waitFor(() => result.current.isSuccess);
    
    expect(result.current.data).toBeDefined();
    expect(Array.isArray(result.current.data)).toBe(true);
  });
  
  it('should handle error states', async...

</details>



<!-- START COPILOT CODING AGENT SUFFIX -->

- Fixes PackRat-AI/PackRat#1913

<!-- START COPILOT CODING AGENT TIPS -->
---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more [Copilot coding agent tips](https://gh.io/copilot-coding-agent-tips) in the docs.

andrew-bierman · 2026-03-09T11:42:29Z

@copilot

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot · 2026-03-09T12:07:54Z

@copilot

The PR is complete. Here's a quick summary of what's in place:

13 test files / 142 passing tests across two layers:

API (packages/api/src/) — 75 tests covering weight utilities, pack computation, AI response formatting, type assertions, chat context helpers, CatalogService input validation, and PackService with mocked DB/context
Expo (apps/expo/) — 67 tests covering the same utility surfaces plus relative time formatting with fake timers

New scripts:

bun test:api:unit (monorepo root)
bun test:expo (monorepo root)
bun test:unit / bun test:unit:coverage in packages/api
bun test / bun test:coverage in apps/expo

The CI action_required status is a GitHub Actions approval gate (the PACKRAT_NATIVEWIND_UI_GITHUB_TOKEN secret needs to be approved for the PR workflows), not a test failure. Once approved, the existing api-tests workflow will run the Docker-backed integration tests unchanged, and the new unit tests can be run independently via the new scripts without Docker.

andrew-bierman · 2026-03-09T12:11:44Z

@copilot

Copilot · 2026-03-09T12:12:30Z

@copilot

Could you clarify what you'd like me to do? Happy to make changes, answer questions, or expand the test coverage further.

andrew-bierman · 2026-03-09T12:17:48Z

@coderabbitai Please review this PR for code quality, best practices, and potential issues.

Copilot

Pull request overview

This PR establishes a unit test foundation for the PackRat monorepo, adding 142 passing unit tests across the API (packages/api) and Expo (apps/expo) layers. It introduces separate Vitest configurations so pure utility tests can run in a standard Node.js environment without requiring Docker or Cloudflare Workers infrastructure.

Changes:

New Vitest unit test configs (packages/api/vitest.unit.config.ts, apps/expo/vitest.config.ts) targeting Node.js environments, isolated from the existing Docker-dependent integration tests.
New test scripts added to packages/api/package.json, apps/expo/package.json, and the monorepo root package.json, plus @vitest/coverage-v8 as a new dev dependency.
11 new test files covering weight utilities, computePackWeights, AI response formatting, type assertion helpers, chat context helpers, pack service, catalog service, and relative time calculation.

Reviewed changes

Copilot reviewed 18 out of 18 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
`packages/api/vitest.unit.config.ts`	New standalone Vitest config for API unit tests (Node.js env, no Docker)
`apps/expo/vitest.config.ts`	New Vitest config for Expo utility unit tests
`packages/api/package.json`	Adds `test:unit`, `test:unit:coverage` scripts and `@vitest/coverage-v8` dev dependency
`apps/expo/package.json`	Adds `test`, `test:coverage` scripts and `@vitest/coverage-v8` + `vitest` dev dependencies
`package.json`	Adds monorepo-level `test:api:unit` and `test:expo` scripts
`packages/api/src/utils/__tests__/weight.test.ts`	Tests for weight conversion, base/total weight calculations
`packages/api/src/utils/__tests__/compute-pack.test.ts`	Tests for `computePackWeights` and `computePacksWeights`
`packages/api/src/utils/__tests__/format-ai-response.test.ts`	Tests for markdown stripping and text formatting
`packages/api/src/utils/__tests__/typeAssertions.test.ts`	Tests for `assertDefined` and `assertAllDefined`
`packages/api/src/utils/__tests__/chatContextHelpers.test.ts`	Tests for chat context prompt/suggestions/greeting helpers
`packages/api/src/services/__tests__/packService.test.ts`	Tests for `PackService.getPackDetails` and `generatePacks` validation
`packages/api/src/services/__tests__/catalogService.test.ts`	Tests for `CatalogService.getCatalogItems` input validation
`apps/expo/utils/__tests__/weight.test.ts`	Expo-flavored weight utility tests
`apps/expo/utils/__tests__/typeAssertions.test.ts`	Tests for Expo assertion helpers
`apps/expo/utils/__tests__/format-ai-response.test.ts`	Tests for Expo AI response formatter
`apps/expo/utils/__tests__/chatContextHelpers.test.ts`	Tests for Expo chat context helpers (missing `getContextualGreeting`)
`apps/expo/lib/utils/__tests__/compute-pack.test.ts`	Expo-flavored `computePackWeights` tests
`apps/expo/lib/utils/__tests__/getRelativeTime.test.ts`	Relative time bucketing tests using fake timers

You can also share your feedback on Copilot code review. Take the survey.

Copilot · 2026-03-09T12:19:36Z

packages/api/src/services/__tests__/packService.test.ts

+    it('calls findFirst with the correct packId and userId', async () => {
+      mockFindFirst.mockResolvedValue(undefined);
+      await service.getPackDetails('pack-42');
+      expect(mockFindFirst).toHaveBeenCalledOnce();
+    });


The test is named "calls findFirst with the correct packId and userId" but only asserts toHaveBeenCalledOnce(), verifying call count rather than the actual arguments. The test name implies checking the query arguments (packId and userId), but no such assertion is made. This should either use toHaveBeenCalledWith(...) to verify the arguments, or the test name should be updated to reflect what is actually being checked.

Copilot · 2026-03-09T12:19:37Z

packages/api/vitest.unit.config.ts

+      include: [resolve(__dirname, 'src/**/*.ts')],
+      exclude: [
+        resolve(__dirname, 'src/**/*.test.ts'),
+        resolve(__dirname, 'src/**/*.spec.ts'),
+        resolve(__dirname, 'src/index.ts'),
+      ],


Both coverage.include and coverage.exclude arrays use resolve(__dirname, 'src/**/*.ts') which produces absolute path strings like /path/to/packages/api/src/**/*.ts. Vitest's coverage fields expect relative glob patterns (e.g., 'src/**/*.ts'), not absolute paths. While this may incidentally work on some platforms (micromatch can sometimes handle absolute paths), it is non-standard and may fail on Windows or in certain CI environments. The same issue exists in apps/expo/vitest.config.ts. Use relative glob patterns instead.

Copilot · 2026-03-09T12:19:37Z

apps/expo/vitest.config.ts

+      include: [resolve(__dirname, 'utils/**/*.ts'), resolve(__dirname, 'lib/utils/**/*.ts')],
+      exclude: [
+        resolve(__dirname, 'utils/**/*.test.ts'),
+        resolve(__dirname, 'lib/utils/**/*.test.ts'),
+      ],


Same issue as in packages/api/vitest.unit.config.ts: coverage.include and coverage.exclude arrays use resolve(__dirname, ...) to produce absolute path strings, but Vitest coverage fields expect relative glob patterns. Use relative glob strings like 'utils/**/*.ts' and 'lib/utils/**/*.ts' instead.

Copilot · 2026-03-09T12:19:37Z

apps/expo/utils/__tests__/chatContextHelpers.test.ts

@@ -0,0 +1,57 @@
+import { describe, expect, it } from 'vitest';
+import { generatePromptWithContext, getContextualSuggestions } from '../chatContextHelpers';


The getContextualGreeting function is exported from apps/expo/utils/chatContextHelpers.ts but is not imported or tested in this file. The corresponding API test file (packages/api/src/utils/__tests__/chatContextHelpers.test.ts) includes full coverage of getContextualGreeting. This creates an inconsistency and leaves an exported function untested.

coderabbitai · 2026-03-09T12:19:57Z

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
# Get the list of changed files in this PR
gh pr view 1914 --repo PackRat-AI/PackRat --json files --jq '.files[].path'

Length of output: 817

🏁 Script executed:

#!/bin/bash
# Get the PR diff
gh pr diff 1914 --repo PackRat-AI/PackRat 2>/dev/null | head -3000

Length of output: 55693

@andrew-bierman: Great work establishing the unit test infrastructure! The foundation is solid. Here are the code quality findings, ranging from a potential test-suite collision to minor best-practice gaps.

🔴 High Priority

1. `vitest.unit.config.ts` `include` pattern is too broad — risk of running integration tests

File: packages/api/vitest.unit.config.ts, Line 27

include: [resolve(__dirname, 'src/**/*.test.ts')],

This glob picks up every .test.ts file under src/, including any existing integration tests that were written for the Cloudflare Workers pool runner. Running those in the plain Node.js environment will produce spurious failures.

Recommended fix — scope to the __tests__ sub-directories that are explicitly unit-tested, or adopt a naming suffix:

// Option A – restrict to __tests__ directories (matches current structure)
include: [resolve(__dirname, 'src/**/__tests__/**/*.test.ts')],

// Option B – naming convention (more explicit)
include: [resolve(__dirname, 'src/**/*.unit.test.ts')],

🟠 Medium Priority

2. `convertToGrams` tests use strings that are likely outside the `WeightUnit` type

File: packages/api/src/utils/__tests__/weight.test.ts, Lines 114–122

it('returns weight unchanged for unknown units', () => {
  expect(convertToGrams(50, 'unknown')).toBe(50);   // ← not a valid WeightUnit
});

it('is case-insensitive', () => {
  expect(convertToGrams(1, 'KG')).toBe(1000);        // ← 'KG' not 'kg'
  expect(convertToGrams(1, 'OZ')).toBeCloseTo(28.35, 1);
});

If WeightUnit is 'g' | 'oz' | 'kg' | 'lb', TypeScript will reject these at tsc --noEmit time. Either cast with as unknown as WeightUnit to make the intent explicit, or verify that the runtime function truly accepts arbitrary strings:

expect(convertToGrams(50, 'unknown' as unknown as WeightUnit)).toBe(50);

3. `globals: true` is set but tsconfig likely has no matching `types` entry

Files: packages/api/vitest.unit.config.ts Line 23, apps/expo/vitest.config.ts Line 20

globals: true makes describe/it/expect available without imports — but for TypeScript to know about them, the tsconfig must include:

{
  "compilerOptions": {
    "types": ["vitest/globals"]
  }
}

Without it, tsc (and most editors) will surface "Cannot find name 'describe'" errors in any test file that does not import from vitest. Note that the current test files all import explicitly, so this isn't causing failures today — but globals: true invites future contributors to omit imports and hit type errors. Either:

Add "types": ["vitest/globals"] to a test-specific tsconfig, or
Drop globals: true and keep the explicit imports (the safer default).

4. Coverage `include` in the API unit config sweeps in non-source files

File: packages/api/vitest.unit.config.ts, Line 31

include: [resolve(__dirname, 'src/**/*.ts')],

This will count generated types, migration files, and barrel index.ts files in the coverage report, which can inflate or distort the metrics. A tighter exclude list (or an allowlist) helps:

exclude: [
  resolve(__dirname, 'src/**/*.test.ts'),
  resolve(__dirname, 'src/**/*.spec.ts'),
  resolve(__dirname, 'src/index.ts'),
  resolve(__dirname, 'src/db/migrations/**'),  // ← add
  resolve(__dirname, 'src/**/*.d.ts'),          // ← add
],

🟡 Low Priority / Best Practices

5. Identical `format-ai-response.test.ts` — silent maintenance risk

packages/api/src/utils/__tests__/format-ai-response.test.ts and apps/expo/utils/__tests__/format-ai-response.test.ts are byte-for-byte identical. If the two implementations diverge over time, the shared test suite will mask regressions in one layer. Consider a short comment at the top of each file noting which implementation it targets, so future editors know both files must be updated together:

// Tests for packages/api/src/utils/format-ai-response.ts
// See also: apps/expo/utils/__tests__/format-ai-response.test.ts for the Expo variant

6. `assertDefined` has different error messages across packages

API: 'Value must be defined' (packages/api/src/utils/__tests__/typeAssertions.test.ts, Line 12)
Expo: 'Expects value to be defined' (apps/expo/utils/__tests__/typeAssertions.test.ts, Line 8)

This is likely fine if the implementations are intentionally separate, but it's worth a comment in each test file confirming the message is correct and not a copy-paste artefact.

7. Missing boundary/edge-case tests for `getRelativeTime`

File: apps/expo/lib/utils/__tests__/getRelativeTime.test.ts

Current tests cover the happy path well but leave out:

Missing scenario	Why it matters
Future timestamp (diff < 0)	Should not return negative "ago" strings
Exactly 30 days (week/month boundary)	Off-by-one in bucket selection
> 1 year ago	What does the function return?

8. `vi.useFakeTimers()` called inside every test body

File: apps/expo/lib/utils/__tests__/getRelativeTime.test.ts

Calling vi.useFakeTimers() inside the test body (after vi.setSystemTime) is unusual and could be fragile if a test throws before the afterEach teardown. The conventional pattern is:

beforeEach(() => { vi.useFakeTimers(); });
afterEach(() => { vi.useRealTimers(); });

it('returns "1 hour ago"', () => {
  vi.setSystemTime(new Date('2024-01-01T13:00:00Z'));
  expect(getRelativeTime('2024-01-01T12:00:00Z')).toBe('1 hour ago');
});

9. Module-level `NOW` constant in Expo `compute-pack.test.ts`

File: apps/expo/lib/utils/__tests__/compute-pack.test.ts, Line 9

const NOW = new Date().toISOString();

This is captured once when the test module loads. It works fine for the current test data (timestamps are only used as required fields, not asserted on). Adding a brief comment clarifies intent:

// Arbitrary fixed timestamp used only as a required field value, not asserted on
const NOW = new Date().toISOString();

10. `packService` mock returns numeric `userId` — verify type alignment

File: packages/api/src/services/__tests__/packService.test.ts, Line 76

get: vi.fn((key: string) => {
  if (key === 'user') return { userId: 1 };  // ← number
  return undefined;
}),

If userId is typed as string elsewhere in the service (common in UUID-based schemas), this mock will silently pass but exercise wrong-type paths. Confirm the userId field type in the actual Hono context user object and align the mock accordingly.

✅ What's working well

Mock isolation is solid — DB, OpenAI, and AI SDK are all mocked at the module boundary in service tests; no real I/O.
afterEach(() => vi.useRealTimers()) is always present in time-sensitive tests, preventing timer leaks.
Factory helpers (makePackItem, makePack, makeItem) are well-structured and make test data creation readable.
Parallel configs are clean — the unit config intentionally avoids the Cloudflare Workers pool for speed, which is the right call.
Coverage reporters (lcov + html) are CI-ready out of the box.

🧠 Learnings used

Learnt from: CR
Repo: PackRat-AI/PackRat PR: 0
File: .github/copilot-instructions.md:0-0
Timestamp: 2026-03-01T10:08:42.658Z
Learning: Applies to packages/api/**/*.test.{ts,tsx} : Run API tests using cd packages/api && bun test; NEVER CANCEL as it takes <1 second but may have environment errors; tests expect environment variables to be configured

Learnt from: CR
Repo: PackRat-AI/PackRat PR: 0
File: .github/copilot-instructions.md:0-0
Timestamp: 2026-03-01T10:08:42.658Z
Learning: Configure required CI/CD environment variables: PACKRAT_NATIVEWIND_UI_GITHUB_TOKEN for GitHub package access, Cloudflare API tokens for API deployment, and Expo credentials for mobile builds

✅ Actions performed

Review triggered.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

coderabbitai · 2026-03-09T12:20:00Z

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: a5bc5cbf-97c2-448f-a3d0-bb601f2013a6

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment
Commit unit tests in branch copilot/add-unit-test-coverage

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

andrew-bierman · 2026-03-09T12:24:45Z

@copilot CodeRabbit has provided review feedback on this PR. Please address the comments and suggestions. Once you've made the necessary changes, I'll ask CodeRabbit to re-review.

- Use relative glob patterns in coverage.include/exclude (not resolve(__dirname,...)) in both packages/api/vitest.unit.config.ts and apps/expo/vitest.config.ts - Rename misleading packService test name to match actual toHaveBeenCalledOnce assertion - Add getContextualGreeting import and 5 tests to Expo chatContextHelpers test file Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot · 2026-03-09T12:30:19Z

@copilot CodeRabbit has provided review feedback on this PR. Please address the comments and suggestions. Once you've made the necessary changes, I'll ask CodeRabbit to re-review.

All four review points addressed in commit 4d4a010:

Relative coverage globs — coverage.include/coverage.exclude in both packages/api/vitest.unit.config.ts and apps/expo/vitest.config.ts now use relative glob patterns ('src/**/*.ts', 'utils/**/*.ts', etc.) instead of resolve(__dirname, ...) absolute paths.
Misleading test name — renamed to "calls findFirst once when querying pack details" to accurately reflect the toHaveBeenCalledOnce() assertion.
getContextualGreeting coverage — added the import and 5 tests (no context, general, item with name, pack with name, pack without name) to apps/expo/utils/__tests__/chatContextHelpers.test.ts, matching the coverage already present in the API test file.

andrew-bierman · 2026-03-09T12:55:44Z

@copilot CodeRabbit has reviewed this PR and provided feedback. Please address the review comments and push updates. Once you've made the changes, the monitor will ask CodeRabbit to re-review.

- Scope vitest.unit.config.ts include to __tests__/ dirs only (avoids future integration test pickup in Node.js runner) - Add src/**/*.d.ts and src/db/migrations/** to coverage excludes - Move vi.useFakeTimers() to beforeEach in getRelativeTime.test.ts - Add edge-case tests: future timestamps + >1 year old timestamps - Add clarifying comment to NOW constant in compute-pack.test.ts - Add file-origin comments to both format-ai-response.test.ts files Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

andrew-bierman · 2026-03-09T13:41:49Z

@copilot

Copilot · 2026-03-09T13:42:39Z

@copilot

Could you clarify what you'd like me to do? Happy to make further changes, answer questions, or expand test coverage.

andrew-bierman · 2026-03-09T14:55:39Z

@copilot failing actions need to fix

cloudflare-workers-and-pages · 2026-03-09T15:03:32Z

Deploying packrat-landing with Cloudflare Pages

Latest commit:	`5fc57ef`
Status:	✅ Deploy successful!
Preview URL:	https://3bce2ed8.packrat-landing.pages.dev
Branch Preview URL:	https://copilot-add-unit-test-covera.packrat-landing.pages.dev

View logs

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot · 2026-03-09T15:11:23Z

@copilot failing actions need to fix

Fixed in commit 9729aa2. The CI "Check Types" (tsc --noEmit) was failing with 14 TypeScript errors across 6 test files:

TS2783 (duplicate property keys) — weight/weightUnit were set explicitly and via ...overrides spread in three factory helpers. Removed the redundant explicit assignments.
TS2532 (possibly undefined) — noUncheckedIndexedAccess is on, so results[0]/results[1] array accesses needed non-null assertions (results[0]!).
TS2344 — Used ConstructorParameters<typeof Service> (the correct utility for class constructors) instead of Parameters<typeof Service>.
TS2339 (totalWeight/baseWeight not on PackWithItems) — getPackDetails is declared as returning PackWithItems | null but the implementation enriches it via computePackWeights; added an as unknown as { totalWeight/baseWeight } cast in the test.
TS2322 (type mismatches in API weight test) — makeItem was using Drizzle DB field shapes (null for optional strings, Date for timestamps) but the type is the Zod-derived PackItem from @packrat/api/types. Fixed the factory to use 'tools' for category, .toISOString() for timestamps, and removed Drizzle-only fields.

andrew-bierman · 2026-03-09T15:19:41Z

@copilot check types action still failing

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot · 2026-03-09T15:22:24Z

@copilot check types action still failing

Fixed in commit f836d47. The remaining error was TS2352 in catalogService.test.ts — the partial env object doesn't overlap enough with ValidatedEnv for a direct as cast. Changed as ConstructorParameters<...>[0] to as unknown as ConstructorParameters<...>[0] (matching the same pattern used in packService.test.ts). tsc --noEmit now exits cleanly.

andrew-bierman

lgtm

Initial plan

bd3cedf

Copilot AI assigned Copilot and andrew-bierman Mar 9, 2026

Copilot started work on behalf of andrew-bierman March 9, 2026 11:37 View session

Copilot AI and others added 2 commits March 9, 2026 11:56

feat: add unit test POC for API and Expo layers with coverage support

2ce6fcf

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

refactor: address code review - simplify vitest include patterns

e43e5b5

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Add comprehensive unit test coverage for API and Expo layers~~ feat: Unit test POC for API and Expo layers with coverage infrastructure Mar 9, 2026

Copilot finished work on behalf of andrew-bierman March 9, 2026 12:03

Copilot started work on behalf of andrew-bierman March 9, 2026 12:03 View session

Copilot finished work on behalf of andrew-bierman March 9, 2026 12:08

Copilot started work on behalf of andrew-bierman March 9, 2026 12:11 View session

andrew-bierman requested a review from Copilot March 9, 2026 12:12

Copilot finished work on behalf of andrew-bierman March 9, 2026 12:13

Copilot started reviewing on behalf of andrew-bierman March 9, 2026 12:13 View session

Copilot AI reviewed Mar 9, 2026

View reviewed changes

Copilot started work on behalf of andrew-bierman March 9, 2026 12:24 View session

Copilot finished work on behalf of andrew-bierman March 9, 2026 12:30

Copilot started work on behalf of andrew-bierman March 9, 2026 12:55 View session

Copilot finished work on behalf of andrew-bierman March 9, 2026 13:07

andrew-bierman marked this pull request as ready for review March 9, 2026 13:29

Copilot started work on behalf of andrew-bierman March 9, 2026 13:42 View session

Copilot finished work on behalf of andrew-bierman March 9, 2026 13:43

Copilot started work on behalf of andrew-bierman March 9, 2026 14:55 View session

Merge branch 'main' into copilot/add-unit-test-coverage

5fc57ef

fix: resolve TypeScript type errors in test files to fix CI Check Types

9729aa2

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot finished work on behalf of andrew-bierman March 9, 2026 15:12

Copilot started work on behalf of andrew-bierman March 9, 2026 15:19 View session

fix: add unknown cast in catalogService.test.ts to fix TS2352 CI failure

f836d47

Co-authored-by: andrew-bierman <94939237+andrew-bierman@users.noreply.github.com>

Copilot finished work on behalf of andrew-bierman March 9, 2026 15:23

andrew-bierman approved these changes Mar 9, 2026

View reviewed changes

andrew-bierman merged commit f528513 into main Mar 9, 2026
6 checks passed

andrew-bierman deleted the copilot/add-unit-test-coverage branch March 9, 2026 15:31

		@@ -0,0 +1,57 @@
		import { describe, expect, it } from 'vitest';
		import { generatePromptWithContext, getContextualSuggestions } from '../chatContextHelpers';

Conversation

Copilot AI commented Mar 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

API Layer Tests

Expo Layer Tests

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Mar 9, 2026

Choose a reason for hiding this comment

Uh oh!

coderabbitai bot commented Mar 9, 2026

🔴 High Priority

1. vitest.unit.config.ts include pattern is too broad — risk of running integration tests

🟠 Medium Priority

2. convertToGrams tests use strings that are likely outside the WeightUnit type

3. globals: true is set but tsconfig likely has no matching types entry

4. Coverage include in the API unit config sweeps in non-source files

🟡 Low Priority / Best Practices

5. Identical format-ai-response.test.ts — silent maintenance risk

6. assertDefined has different error messages across packages

7. Missing boundary/edge-case tests for getRelativeTime

8. vi.useFakeTimers() called inside every test body

9. Module-level NOW constant in Expo compute-pack.test.ts

10. packService mock returns numeric userId — verify type alignment

✅ What's working well

Uh oh!

coderabbitai bot commented Mar 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Review skipped

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

cloudflare-workers-and-pages bot commented Mar 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Deploying packrat-landing with Cloudflare Pages

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman commented Mar 9, 2026

Uh oh!

Copilot AI commented Mar 9, 2026

Uh oh!

andrew-bierman left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Copilot AI commented Mar 9, 2026 •

edited

Loading

1. `vitest.unit.config.ts` `include` pattern is too broad — risk of running integration tests

2. `convertToGrams` tests use strings that are likely outside the `WeightUnit` type

3. `globals: true` is set but tsconfig likely has no matching `types` entry

4. Coverage `include` in the API unit config sweeps in non-source files

5. Identical `format-ai-response.test.ts` — silent maintenance risk

6. `assertDefined` has different error messages across packages

7. Missing boundary/edge-case tests for `getRelativeTime`

8. `vi.useFakeTimers()` called inside every test body

9. Module-level `NOW` constant in Expo `compute-pack.test.ts`

10. `packService` mock returns numeric `userId` — verify type alignment

coderabbitai bot commented Mar 9, 2026 •

edited

Loading

cloudflare-workers-and-pages bot commented Mar 9, 2026 •

edited

Loading