Edge-case tests
Write tests for {{function_or_feature}} covering: normal cases, boundaries, invalid input, concurrency and failure of dependencies. Each test name should describe the behaviour. Do not change the implementation.
Operations
Best agent for testing & qa
Writing tests, exploratory testing and catching regressions.
Full ranking
Task index, 0 to 100. Bars are scaled to the field, so small gaps look small.
- 1
- 2
- 3
- 4
Key takeaways
- 1Agents write good unit and integration tests when given clear behaviour specs.
- 2Computer-use agents can now do exploratory UI testing.
- 3Watch for tests that assert the bug.
Recommended setup
Coding agent generating tests in CI, plus a computer-use agent running exploratory passes on staging.