测试用例生成Generate test cases

从需求生成测试用例与边界条件。 输入：需求说明、输入域、已有测试。

Output a test-case table covering normal, boundary and error inputs; do not claim execution. Inputs: Requirements, input domain, and existing tests.

为需求 {requirement} 生成测试用例。输出 cases[]：id、type(normal/boundary/error)、input、expected、rationale。expected 必须能从需求推出；需求未定义的行为标 undefined_behavior，不猜。列出 coverage_gaps[]。

占位符用 {字段} 标注，运行前替换为本次输入。

Generate test cases for {requirement}. Output cases[] with id, type (normal/boundary/error), input, expected and rationale. expected must follow from the requirement; mark undefined behaviour instead of guessing. List coverage_gaps[].

Replace {field} placeholders with the current inputs before running.

cases[] 与 coverage_gaps[]。

cases[] and coverage_gaps[].

• 三类 type 都有或说明缺因。
• expected 可回连需求。
• 未定义行为不填猜测。

• All three types exist or gaps are explained.
• Expected results link to requirements.
• Undefined behaviour is not guessed.

• 用例重复：按输入域合并。
• 需求模糊：标 undefined_behavior。

• Duplicate cases: merge by input domain.
• Vague requirement: mark undefined_behavior.

提示词效果未实测。机制依据见出处；换模型、版本或输入会改变结果。

This site has not tested the effect of this prompt. See the sources for the mechanism; changing model, version or input can change the result.