Unit Test Generator with Real Edge Cases

Generates unit tests that actually probe boundaries and failure modes — not three happy-path tests with different variable names.

Fill in the blanks

Prompt variables

3 variables detected — fill them and Copy inserts your values (0/3 filled).

Write unit tests for the code below. Quality bar: each test must be able to fail for a distinct reason. If two tests can only fail together, merge them. Cover, in this order: 1. **Happy path** — one or two representative cases, no more. 2. **Boundaries** — empty, single element, exactly-at-limit, one-past-limit, zero, negative, maximum values. 3. **Invalid input** — nulls, wrong formats, out-of-range; assert the specific exception/error, not just "throws". 4. **State & interaction edges** — called twice, called concurrently (if relevant), dependency returns error, dependency times out. 5. **The nasty ones** — unicode and whitespace in strings, DST transitions and leap days in dates, floating point near-equality, mutation of input arguments. Conventions: - Test names describe behavior: `rejects_expired_token_with_401`, not `test3`. - Arrange / Act / Assert with a blank line between sections; no logic in tests. - One behavioral assertion per test (multiple asserts fine if they verify one behavior). - Use {{framework}} and follow its idioms. - If some part of the code is untestable as written, write the tests you can, then note exactly what refactor would unlock the rest. Finish with a coverage gaps section: what you still couldn't test and what risk that leaves. ```{{language}} {{code}} ```
Field notes

Comments (0)

LLM Atlas

A free map of prompts that actually work — tested on real models, starred by real people.

© 2026 LLM AtlasPrompts default to CC0 unless stated.