Knowledge documents in Testing.
How to pick a base point when you follow the import graph backward from changed files to run only the related tests, the fail-closed rule that falls back to a full run when things are uncertain, and why the final gate is kept separate
How to read set, logic, and quantifier notation in specifications and proofs by mapping it 1:1 to TypeScript code and Bend syntax, and the test gaps caused by misreading symbols such as vacuous truth, counterexamples, and reverse implication
How to move boundary value analysis from runtime tests to the type level, pairing a passing and a rejecting witness on each of the 7 extremes of the type lattice — and the axes types can never cover, like the time axis and side-effect counts
How to verify compilation instead of execution so that 'what wrong code does this type actually block' is pinned down in code, plus the @ts-expect-error and mutation checks that tell you whether the test is still alive
Planting artificial defects in code to see whether the existing tests fail, exposing the empty assertions and loose checks that coverage hides
The problem of deciding whether a result is right or wrong, and why tests whose expected values were reverse-engineered from the implementation verify nothing
Old-browser compatibility regressions can't be caught by a render-only smoke — you must run real click/submit interactions and promote runtime errors via a broad console-error filter
How to separate flaky from deterministic failed tests in E2E CI using retry results, and distinguish timing races from build defects
How to record per-id render counts with a jest.mock probe and compare against a baseline snapshot to lock React re-render isolation into a unit test
A pattern for stably importing an ESM-only npm package in a Jest CJS environment via a vendor TS file and moduleNameMapper