tasclass — TAF coding course

Maps to: SYSTEM.md Phase 1 (app-research) Goal: leave this lesson with one clear paragraph describing an app you're actually going to build.

Why this phase exists

Every app in this system starts here, whether it's built in Swift or (like this course) React Native — market/idea validation doesn't depend on what code the app is written in. SYSTEM.md Phase 1 shares the exact same app-research skill as the native Swift system for this reason: a good app idea is a good app idea regardless of stack.

What makes a good beginner app idea

A real frustration, not a hypothetical. "I keep forgetting to drink water" beats "an app for productivity" every time — specificity is what makes the next 8 lessons concrete instead of vague.
One core feature, not five. You will build ONE thing that works well. Everything else is a "maybe later."
No account system required. Login/signup adds real complexity (Phase 7's account-deletion requirement, for one) that isn't worth it for a first app. Local-only apps are completely valid.

Worked example: the Habit Tracker

This course demonstrates every lesson on one running example: a Daily Habit Tracker. The one-paragraph concept:

"An app where I can add habits I'm trying to build (like 'drink water' or 'read 10 minutes'), and each day I tap a habit to mark it done. The app shows a streak count for each habit so I can see how many days in a row I've kept it up. No accounts, no login — just habits and streaks, saved on my phone."

Notice what this paragraph does: it names the ONE core feature (mark a habit done, see a streak), explicitly rules out scope creep (no accounts), and describes something a real person would actually use. Every lesson from here on builds this exact app, one phase at a time.

Validate the idea with Claude Code

Ask Claude Code to use the app-research skill (it's shared with this system's native-Swift portfolio — see docs/course/05-skills-reference.md Part 3) to check: are there similar apps already? What do they do well or poorly? This isn't about avoiding competition — plenty of great apps have direct competitors — it's about knowing what you're building relative to.

Try this prompt:

I want to build a simple daily habit tracker — add habits, tap to mark done each day, see a streak count. No accounts. Use the app-research skill to check what similar apps exist and what I should know before I start.

Deliverable

One clear one-paragraph app concept — the exact deliverable 04-workflow-course-overview.md Module 1 asks for. If you're following along with the Habit Tracker, you already have it (above). If you're building your own idea, write your paragraph now, in the same shape as the example: what it does, what the ONE core feature is, and what you're explicitly leaving out.

Apply to your own idea

Write your own one-paragraph concept using the same three rules above (real frustration, one core feature, no account system needed). Run the same app-research prompt with your idea substituted in. Don't move to Lesson 2 until you have this paragraph written down somewhere you can copy from — you'll need it verbatim when you scaffold your project.

Prompts

Maps to: courses/react-native-course/lessons/01-idea-and-research.md (SYSTEM.md Phase 1) Run this: at the very start, with the student in the room, before anything else.

Prompt 1 — Capture the idea (run this yourself, on your phone)

Open ChatGPT or the Claude app on your phone — the free version is fine — and paste this in, filling in your own idea:

I want to build my first app. Here's my idea: {{DESCRIBE YOUR APP IDEA IN YOUR OWN WORDS}}

Help me turn this into a clear one-paragraph app concept: name ONE core feature the app absolutely needs, and tell me what to leave out for a first version. Ask me questions before you lock in the concept — don't just repeat back what I said.

Keep going back and forth until the concept feels right — better to spend a few extra minutes here than build the wrong thing for two weeks.

Prompt 2 — Full app plan (run this in ChatGPT on your phone)

Once the concept from Prompt 1 is solid, paste this into ChatGPT — this is where the real planning happens, before any building starts:

I'm planning an app called {{APP_NAME}}. Here's the concept: {{PASTE concept.md HERE}}

We're building this with Expo Router + TypeScript + Zustand. Write me a full build plan:
1. The screens/pages the app needs, one by one, and what each one does
2. The data the app needs to remember (what gets saved, in what shape)
3. The exact order to build things in, step by step
4. Anything from the stack above I'll specifically need to set up for this app
5. How I'll know each step is actually done (a quick test I can do myself)

Write the whole thing as a clean markdown document I can save and come back to.

Save ChatGPT's response — this becomes {{APP_NAME}}'s planning doc. Paste it into your dashboard on the class website (tasclass.online) once you're logged in, so it's tracked and your instructor can review it before you start building.

After this prompt

Save the student's own words (their original idea, as close to verbatim as they wrote it) to student-apps/<slug>/transcript.md.
Save the one-paragraph concept the AI helped land on to student-apps/<slug>/concept.md.
See student-apps/README.md at the repo root for the exact naming convention (<slug> = student's first name once known, or student-N as a placeholder, or the app name if that comes first) — rename via git mv once the real name is settled, don't duplicate the folder.

Full lesson

For the "why" behind this phase, see courses/react-native-course/lessons/01-idea-and-research.md.

Lesson 1 — Idea & Research