Testing & quality

Software test coverage

This page shows the coverage report for a Next.js + TypeScript feature I worked on. The goal was to make changes safe to ship by relying on strong automated tests rather than manual checks.

Coverage report showing high percentages across statements, branches, functions, and lines
Coverage summary for a production feature module.

Coverage snapshot

Every pull request must keep coverage above the guardrails below. The dashboard lives in a Next.js monorepo, so enforcing this baseline protects dozens of shared components and hooks.

Statements

97.9%

Branches

91.3%

Functions

91.2%

Lines

97.9%

These numbers mean almost every branch—loading, empty, slow, error—is executed by automated tests before a release.

How I enforce it

  • Unit tests for shared utilities, selectors, and hooks using Jest + ts-jest.
  • Component tests with React Testing Library to lock down pagination, filters, and debounced search.
  • MSW API mocks stress slow responses, 500s, and malformed JSON so the UI never regresses on edge cases.
  • GitHub Actions blocks merges when coverage drops and leaves inline summaries on each PR.
  • Nightly Playwright smoke suite runs against preview builds to catch real-browser issues.

Command used locally: npx jest --runInBand --coverage --coverageThreshold '{ "global": { "statements": 0.95, "branches": 0.85 } }'

Impact

  • Refactors of the reporting table shipped with zero regressions across six releases.
  • Coverage gate prevented three bugs from reaching production within the first month.
  • New engineers can ship confidently because baseline scenarios are already codified.
← Back to portfolio