Test Loop

Run the test suite, analyze failures, fix them, and repeat until all tests pass.

**Step 1: Run tests**
```bash
PYTHONPATH=backend pytest backend/tests/ -v --tb=short 2>&1 | head -150

Step 2: For each failing test

• Read the test to understand what it expects
• Read the implementation code being tested
• Identify the root cause: implementation bug vs outdated test
• Fix the implementation if it is a real bug
• Update the test only if behavior has intentionally changed — confirm with me first

Step 3: Re-run Run only the previously failing test to confirm fix, then run the full suite.

Step 4: Report When all tests pass, summarize:

• How many tests were failing at start
• Root causes found (bug / stale test / missing fixture / import error)
• What was changed

Do not stop until all tests pass or you hit a blocker that requires my decision. ```