Prep RoomSign inStart free
Code RoomTest generation for pure functions
MediumPrep Room Coding #4318

Test generation for pure functions

Vibe & agenticAlgorithms & data structuresMid–Senior~15 min

Your team wants ~80% unit-test coverage added to a stable, well-typed Python utility module (pure functions, no I/O). One engineer sets up an autonomous agent to 'add tests until coverage hits 80%, running pytest in a loop.' Another just one-shots the whole test file from the module source. Given the module's nature, which fits — and what's the actual risk you're managing?

Implement
evaluate_test_suite(covered_lines: int, total_lines: int, test_count: int, asserting_test_count: int, target_percent: int) → str
Examples
in[80,100,12,12,80]out"ready"
in[95,100,20,17,80]out"gamed_metric"
in[60,100,10,10,80]out"below_target"
What a strong answer looks like

Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.

0:00 of about 15 min

Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.

Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.