Capstone: Task Tracker Pro
This capstone builds Task Tracker Pro, a small Rails task tracker with one Task model, test first at every step. Each feature starts with a spec that runs red for a specific reason, followed by the smallest code that turns it green. The course moves through model, request and system specs with RSpec, FactoryBot and Capybara. It closes a missing Capybara gap in the app and reviews an AI-suggested feature against a failing test. The full suite ends at twelve passing examples.
What you'll learn
- ✓Writing a spec first and watching it fail for a specific reason
- ✓Making the smallest migration, model and factory change that turns a spec green
- ✓Driving routes and a controller from a request spec
- ✓Writing a Capybara system spec with the rack_test driver and the views it needs
- ✓Reviewing an AI-suggested change with evidence and a failing test
- ✓Running a model, request and system suite together
Planning: one model, TDD from the first line
The ninety-six junior courses before this one each taught a piece: ActiveRecord depth, controller mechanics, RSpec's toolkit, Git workflows, code review, AI used with care. None of them, on its own, built a working, fully tested app end to end. That's this capstone's job, across seven lessons. It…
Read the full lesson free →Curriculum
7 lessonsPlanning: one model, TDD from the first line
Defines the Task model, its four actions and the test-first order every later lesson follows.
Red: a model spec for a model that doesn't exist yet
Writes the Task model spec and runs it red, failing with NameError because nothing exists yet.
Green: the migration and model that make it pass
Adds the smallest migration, model and factory that make all six model examples pass unchanged.
Red, then green: a request spec before the controller
Writes a request spec before any route exists, then adds routes, controller and minimal views to pass it.
A system spec, and a gap this app had
Adds Capybara, sets the rack_test driver inside before, and builds the views a system spec needs.
An AI-assisted feature, reviewed like any other diff
Catches an AI sorting suggestion that orders enum priorities backwards, using a review comment and a failing spec.
Closing the arc: from red to a tested, AI-assisted app
Runs all twelve examples across model, request and system layers and reviews what each lesson added.