ia-writing-tests

To write tests that verify software behavior, enforce one-behavior-per-test structures with outcome-based assertions and derive test cases from user requirements, implemented features, and stated claims.

A silent failure in test writing occurs when errors are swallowed without breaking the test. You cover it by enforcing explicit assertions on error paths and observable outcomes during test generation.

Avoid test anti-patterns like mock-overuse and snapshot-only testing by applying mocking discipline and ensuring each test asserts observable outcomes rather than relying on circular assertions or snapshots.

Derive test cases from requirements and implemented features by cross-referencing user requirements, software diffs, and response claims, ensuring explicit coverage for both edge cases and silent failures.

Yes, you can apply these test writing rules across any language or framework, because the discipline focuses on outcome-based assertions and anti-pattern avoidance rather than framework-specific implementations.

Tests pass but miss edge cases and broken outcomes when they lack coverage built from requirements, implemented behavior, and stated claims, creating test theater instead of real behavior verification.