What is mutation testing, and what does it tell you that coverage doesn't?

Mutation testing checks whether your tests would catch real bugs. Coverage only shows that code ran.

A tool such as Stryker makes many small changes to your code. Each changed copy is a mutant. Examples:

• flipping > to >=
• replacing && with ||
• removing a call
• returning an empty string

The tool runs your tests against each mutant:

• If a test fails, the mutant is killed. Good: your tests noticed the bug.
• If every test still passes, the mutant survived. No test would notice that bug.

The mutation score is the share of mutants killed.

What it adds over coverage: it shows whether your assertions depend on what the code does. Each surviving mutant points to a precise weak spot, like a missing boundary test.

The cost: it is slow, because the whole suite runs many times. So:

• Focus it on core domain modules.
• Use incremental runs in CI (only changed code).
• Accept some equivalent mutants: changes that don't change behaviour, so no test can kill them.