PromptSharp › Prompt Library › Dev & Engineering › Edge-case hunt: the failure inputs your happy-path tests will miss

Edge-case hunt: the failure inputs your happy-path tests will miss

Your tests pass but you don't trust them. Enumerate the boundary and failure cases that the happy path never touches.

The prompt — copy and run it

You are a test engineer building an edge-case and failure-mode inventory for a function or feature I'll implement.

Produce:

A) BEHAVIOR RESTATEMENT — the contract in plain English: inputs, outputs, and what 'correct' means, from my description.

B) EDGE-CASE TABLE — boundary values, empty/null, oversized, malformed, concurrency, and failure-injection cases, each with the input, the expected behavior, and why it matters.

C) MISSING-COVERAGE READ — given the tests I pasted (if any), which of the above are untested.

D) PRIORITY — the 3 cases most likely to bite in production, to write first.

Inputs: [WHAT THE CODE DOES] · [SIGNATURE / INPUTS + TYPES] · [EXISTING TESTS, IF ANY] · [WHERE IT RUNS]

Rules: Do not assume behavior I didn't specify — where the contract is ambiguous, list it as a question so I define it. Don't invent framework APIs. Keep proprietary code out of consumer AI tools. This enumerates cases; writing and verifying the tests stays yours. Do not invent facts, numbers, or details you weren't given.

Why this prompt works

Green test suites give false confidence because they only cover the inputs the author imagined; a structured sweep across boundary, null, malformed, and concurrency classes plus a coverage-gap read against existing tests surfaces the failure inputs the happy path never touches — and routing ambiguity to questions forces the contract to get defined rather than assumed.

Frequently asked

Related Dev & Engineering prompts