List top errors
This command lists the run’s errors, most affected tests first. Use this when a run has many failures and you want to find the shared causes. In pretty output it prints a block per error rather than a table.<run_id>- the run whose errors to group, fromruns list.--project <project_id>- the project the run belongs to.
Sample output (--json)
test_entity_ids names every test in the group. test_entities describes a capped subset of them, so on a large group the two differ in length - read any test beyond the described ones with test-entities show (get still works as an alias). The pretty format makes this explicit by printing a +N more line with the remaining ids.
The error log is the grouping key with Playwright’s per-test Call log: section stripped, which is what lets tests that failed the same way land in the same group.
The response is not paginated and returns at most 20 groups.