Feature Request: read trace for automatically debug

[hikouki-gumo]

Description

When running E2E test with Playwright, test may failed for many reasons. QA team has to run test manually, read trace viewer to debug, update code and loop process until fix all. It may boost QA team performance if this process is automatic.

Use case

Automatically debug Playwright test.

1. Input: Current version of Playwright code (or better, natural language test plan).
2. Action: Loop until pass all tests

• Agent run tests
• If test failed:
  • Read trace
  • Update Playwright code

3. Output: New version of Playwright code that pass all the tests.