run-tests

name	run-tests
description	Run project tests. Use after modifying code to catch regressions before committing.
disable-model-invocation	true
allowed-tools	Bash, Read, Grep
argument-hint	[backend|frontend|mobile|all] [--file path] [--app appname]

Run Tests

Run the test suite for the project. Default runs all layers.

Usage

• /run-tests — Run all layers
• /run-tests backend — Django backend tests
• /run-tests frontend — React frontend tests
• /run-tests mobile — React Native (RNTL) mobile tests
• /run-tests backend --app accounts — Specific Django app
• /run-tests backend --file tests/ai/test_service.py — Specific test file
• /run-tests frontend --file src/__tests__/Component.test.tsx — Specific test file

Commands by Layer

Backend (pytest)

cd backend && pytest --tb=short
# Specific app:
cd backend && pytest tests/<app>/ --tb=short
# Specific file:
cd backend && pytest <path> -v
# With coverage:
cd backend && pytest --cov=. --cov-report=term-missing

Frontend (Jest + RTL)

cd frontend && npx jest --no-coverage
# Specific file:
cd frontend && npx jest --no-coverage <path>
# Watch mode:
cd frontend && npx jest --watch

Mobile (RNTL)

cd <mobile-dir> && npx jest --no-coverage
# Specific file:
cd <mobile-dir> && npx jest --no-coverage <path>

Type Check (no test runner needed)

cd frontend && npx tsc --noEmit

Linting

# Backend
cd backend && ruff check .
# Frontend
cd frontend && npm run lint

Steps

1. Identify which layer to test based on $ARGUMENTS (default: all)
2. Run the tests
3. Analyze output:
   • All passed → report success with count
   • Failures → list failed tests, analyze if pre-existing or new
4. For new failures → invoke systematic-debugging

Output Format

Success:

✓ Backend: 87 passed in 4.2s
✓ Frontend: 23 passed in 8.1s
✓ TypeScript: no errors

Failure:

✗ Backend: 2 failed, 85 passed
  FAILED tests/autodev/test_pipeline.py::test_spec_generation - AssertionError
  FAILED tests/ai/test_service.py::test_litellm_fallback - ConnectionError

→ Recommend: run systematic-debugging on these failures

Known Pre-Existing Failures

Document any known flaky tests here when discovered:

# Add entries as: test_name — reason — since version