Skip to content

kaininx/PlainTab / .agents/skills/tdd

Use when implementing PlainTab features or bug fixes with test-first development, red-green-refactor, regression scripts, browser checks, or behavior-focused validation before production code changes.

PlainTab TDD

Use this skill when changing runtime behavior. Adapted for PlainTab from Matt Pocock's tdd skill.

Philosophy

Tests should describe observable behavior through the public interface, not internal structure. In PlainTab, the public interface might be a runtime module API, a settings interaction, a stored config shape, a wallpaper source lifecycle, or the rendered new-tab page.

PlainTab has no package manager or test framework. Do not add one. Use focused Node scripts, browser automation, and existing validation commands.

Before Code

  1. Read .claude/rules/00-core.md and the relevant module rule.
  2. Identify the behavior to lock down in user-facing language.
  3. Choose the smallest useful feedback loop:
    • Node script in docs/ai-tasks/ for pure JS behavior.
    • Browser/Playwright check for DOM, settings, command palette, or wallpaper rendering.
    • Manual web-mode check only when automation is not practical.
  4. Name the check after behavior, not implementation.

Red-Green Loop

Work one vertical slice at a time:

  1. RED: write one failing check for one behavior.
  2. Verify RED: run it and confirm it fails for the expected reason.
  3. GREEN: write the smallest production change to pass.
  4. Verify GREEN: run the check again.
  5. Repeat for the next behavior.

Do not write a batch of tests and then a batch of implementation. Do not refactor while RED.

PlainTab Test Shape

Good checks:

  • Exercise real modules and real serialized config shapes.
  • Preserve existing localStorage keys unless the task includes a migration.
  • Cover extension/web-mode differences when the behavior depends on browser APIs.
  • Assert visible wallpaper safety when touching image layers: at least one layer should remain visible.
  • Assert language key completeness when adding UI copy.

Avoid:

  • Adding npm, package.json, lint frameworks, test frameworks, or frontend frameworks.
  • Tests that only assert private helper names or DOM implementation details unrelated to behavior.
  • Long startup work in repro code that would never exist in the real first-paint path.

Validation

Use the narrowest meaningful command, then broaden when the change is shared:

Get-ChildItem -Recurse js -Include *.js | ForEach-Object { node --check $_.FullName }

For i18n changes, use the update-i18n skill. For visual or interaction changes, verify in a real browser whenever practical.

finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 319kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 319kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 319kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 196kUpdated Claude CodeCodexVerifiedInstall