# Finding Template

## Title

Short, falsifiable observation.

## Status

`proposed | reproduced | confirmed | disputed | resolved`

## Summary

One paragraph.

## Hypothesis

What did we expect to happen?

## Test Setup

- Project version:
- Classifier version:
- Test URL:
- Prompt or triggering action:
- Timestamp window:
- Environment:

## Raw Evidence

Include sanitized request records:

```json
{
  "timestamp": "2026-06-23T00:00:00.000Z",
  "eventType": "server_page",
  "path": "/lab/root",
  "ipHash": "optional-for-public-report",
  "userAgent": "...",
  "headers": {}
}
```

## Expected Result

What should a crawler/browser/human have done?

## Observed Result

What actually happened?

## Reproduction Steps

1. Deploy commit SHA.
2. Clear or mark dashboard start time.
3. Trigger the exact prompt or crawl.
4. Wait the defined time window.
5. Export matching raw events.
6. Compare to expected result.

## Interpretation

Explain why this evidence supports the finding.

## Limitations

False positives, missing data, platform caveats.

## Proposed Next Test

What would confirm or falsify this more strongly?