Skip to content

Test Quality

Having tests isn't enough — they need to be good tests. The /twd:test-quality skill analyzes your test files and grades them across four dimensions, giving you actionable feedback on where to improve.

Running the Skill

plaintext
/twd:test-quality

The skill reads each of your TWD test files and evaluates them against four weighted criteria:

  • Journey Coverage (35%) — Do tests cover complete user workflows, not just visibility checks?
  • Interaction Depth (20%) — Is there variety in the user actions being tested?
  • Assertion Quality (25%) — Are assertions verifying real outcomes (payload checks, state) vs. loose checks?
  • Error & Edge Cases (20%) — Are failure scenarios and boundary conditions covered?

Each file gets a letter grade (A through D) with a weighted overall score:

TWD Quality Grades - Overview

TWD Quality Grades - Detailed breakdown

Improving Your Tests

The skill provides 2-3 actionable suggestions per file that scores below A, referencing the actual test content. You can then run /twd again to implement the improvements:

What's Next

Your tests are written, running in CI, covering your routes, and scoring well. The final step: generate a visual gallery so everyone on the team can see what's being tested.

Released under the MIT License.