Project ideas · Graded by rung · Live online · Worldwide

Twelve vibe coding projects that actually get finished.

The internet has no shortage of project lists. What it lacks is honesty about why student projects die: they are chosen too large, they hide a difficulty nobody warned about, and they have no finish line. This page fixes all three. Twelve ideas graded across four rungs, from a first afternoon to a portfolio piece, each with the trap inside it named in advance, because the trap is the education. Then the 21-day plan we use in class to walk a student from idea to shipped, and what reviewers of every kind actually look for when they meet the result.

Live online · ages 6 to 67 · taught from India, worldwide · from USD 100 a month · first class free

Start here

The project-first courses behind this page

Every idea below is the kind of thing these courses finish weekly, with a teacher doing the scoping and a batch doing the witnessing.

The short answer

Good vibe coding projects for students share three properties: they are small enough to finish, they solve an irritation the student personally feels, and they contain one honest difficulty that forces learning, like persistent state, a live data source, or two users wanting the same thing. This page grades twelve such ideas across four rungs from first afternoon to portfolio piece, names the trap inside each, and lays out the 21-day idea-to-shipped plan used in Modern Age Coders classes. Reviewers of finished projects reward four things: it runs live, a note explains one decision, someone real has used it, and the student can tour the code cold. Classes run worldwide at USD 100 a month for groups, USD 150 one to one, first class free.

The ladder

Twelve ideas, four rungs, every trap named

Reaim any of these at your own life freely; the rung and the trap are what matter, not the noun. The trap is not a warning, it is the syllabus.

Rung one: a first afternoon

The pocket-money ledger. Three buttons: earned, spent, balance. The trap is money itself: decimals that drift and spending more than you have, which is to say validation, the first real concept. The chore rotator. Whose turn is it, across three siblings and five jobs? The trap is fairness over weeks, which quietly requires state that survives closing the tab. The homework countdown board. Every deadline, sorted, with days remaining. The trap is dates: time zones, midnight, and what "3 days left" means at 11 p.m.

Rung two: a weekend or two

The bus-stop mirror. One page that answers "coat or no coat" from a public weather source. The trap: the source will be down one morning, and the page must say so instead of lying. The quiz-night scoreboard. Teams, points, a winner. The trap is the disputed point and the tie, the first taste of rules as code. The stubborn flashcards. Cards you get wrong return sooner. The trap is that scheduling logic is real logic, and the model will get it subtly wrong first try.

Rung three: the 21-day term project

The league table. Fixtures in, standings out, for any local competition. The trap is ordering: points, then goal difference, then what? Real tables have tiebreak rules and now the student does too. The club signup sheet. Twenty places, first come first served. The trap is two people taking the last place at once, the gentlest possible introduction to a race condition. The reading log. Books, ratings, covers fetched from a public book API. The trap is other people's data: missing covers, three editions, absent authors.

Rung four: the portfolio piece

The revision planner. Exam dates in, a day-by-day plan out. The trap is constraints that clash, when there are more topics than days. The recycling reminder. What goes out tonight, per your area's actual rules. The trap is that civic data is inconsistent, and handling mess gracefully is a senior skill. The commissioned tool. Built for one real user who is not you, to their requirements. The trap, and the crown: they will want something you would never have guessed, and you have to ask.

The difference between dying and shipping

The same idea, before and after scoping

Here is the single conversation that decides a project's fate, shown honestly. It happens in week nine of our teen track and in the first class for anyone who arrives with an idea.

AS THE STUDENT PITCHES ITthe version that dies
"A fitness app. You log workouts,
 it makes graphs, there are streaks
 and badges, you can follow friends,
 there's a leaderboard, maybe an AI
 coach that suggests workouts, and
 it works on phones as an app."

# Seven features. Four systems.
# Accounts, social, notifications,
# app stores. Estimated honestly:
# four months. Available: three weeks.

Nothing is wrong with this dream. As a first scope it has a survival probability of roughly zero, and every abandoned-project pile is made of exactly these.

AFTER THE SCOPING CONVERSATIONthe version that ships
"MY pushup tracker. One user: me.
 Each day: date, count, ten seconds
 to log. One chart: last 30 days.
 One streak number. Works in my
 phone's browser. Nothing else.

 v2 IF v1 survives two weeks of
 real use: my brother gets one."

# One feature done well, one chart,
# one real user, three weeks, ships.

Same dream, first slice. The student who ships this learns more than the student who abandons the grand version, and, crucially, still wants to build v2.

The scoping move is always the same three questions. Who is the one user this must work for? What is the one action they do daily? What can be deleted and added back later? Vibe coding makes starting so cheap that scoping is now the scarce skill: the model will happily scaffold the four-month version on day one, and the wreckage looks impressive right up until it is abandoned.

What reviewers reward at the far end has nothing to do with size either. A project that runs live at a link, a short note explaining one decision and one failure, evidence one real person used it, and a builder who can tour the code cold: those four signals beat scale and polish in school showcases, hackathons and interviews alike, and every rung on this page's ladder is designed so all four are reachable.

The plan

Idea to shipped in 21 days

The rhythm we run rung-three projects on, in class and out. Evenings, not heroics, and every phase ends with something that runs.

  1. Days 1 to 2: cut the scope, write the spec

    The three scoping questions, answered in writing: one user, one daily action, everything else deleted. The spec is ten lines in the style this page's right-hand panel shows, and it is the contract the student checks every prompt against for three weeks.

  2. Days 3 to 7: the ugly walking skeleton

    Every part exists and connects, nothing is pretty, the trap has been met. The league table sorts wrongly, the API fails rudely, and that is on schedule: meeting the trap early is the whole reason the trap was named in advance.

  3. Days 8 to 12: defeat the trap properly

    The week the project earns its place in a portfolio. The tiebreak rule gets written, the down API gets a polite fallback, the race for the last club place gets settled. This is also the week the student reads more code than they prompt, on purpose.

  4. Days 13 to 16: one real user, watched silently

    The hardest exercise we set: hand it over and say nothing. Where the user hesitates is the to-do list. Students discover their labels are riddles and their obvious button is invisible, and no amount of teacher feedback substitutes for this fifteen minutes.

  5. Days 17 to 19: fix what the watching found, then stop

    Fix the top three findings only. The rest go in the note as known limits, which reviewers respect far more than they respect the word "coming soon". Feature-adding past day 19 is how day-30 projects are born, and day-30 projects are how abandonment starts.

  6. Days 20 to 21: ship, write the note, tell somebody

    A live link, a two-paragraph note covering one decision and one thing that went wrong, and an audience of at least one. In class this is demo day; at home it can be the family group chat. Shipping to silence is only half a finish.

The catalogue

Courses that run on projects, by age

Project-based is not a format we bolt on; it is how every track here already works. These are the ones students on this page choose.

I

Project-first tracks

Chosen for this page rather than listed in full

SHIP / 01

Vibe Coding for Kids

Rung-one projects at kid scale: games and gadgets finished weekly in Scratch, with the finishing habit installed early.

Open the syllabus

SHIP / 02

Vibe Coding for Teens

The ladder's main home: rungs one to three across a term, ending in a scoped chosen project and a demo day.

Open the syllabus

SHIP / 03

Hackathon Prep for Teens

Projects under judgement: timeboxed builds, honest demos and pitches, for students who want competition stakes.

Open the syllabus

SHIP / 04

Vibe Coding for College

Rung four as a way of life: deployed projects with real users, built to withstand interview questioning.

Open the syllabus

Students shipping from this page also take Full-Stack Web for Teens · App Development for Teens · JavaScript for Teens · Git and GitHub for Teens · Data Science for Teens · Python for Teens · Hackathon Prep for Kids. The full set is on the catalogue.

Fees

What it costs, anywhere in the world

Project teaching costs what all our teaching costs. The figures are flat across every country we serve outside India, and a family can leave at any month end without a retention call.

Free first class

USD 0

no card required

  • A real class, not a sales call
  • Doubles as a placement check
  • You watch it, then decide
Book it

Group batch

USD 100

a month, billed in US dollars

  • Five to eight students, same teacher every week
  • Live video, never a recording
  • Build work between sessions, reviewed by the teacher
  • Certificate on completion
Start here

One to one

USD 150

a month, billed in US dollars

  • Private teaching, at the pace the goal needs
  • Scheduling built around your family
  • Syllabus shaped to one objective
Enquire

What learners and families say

Rated 4.9 across 547 Google reviews

Real reviews from real families. We neither write nor commission them.

★★★★★

"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."

Ria Mukherjee

Parent

★★★★★

"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."

Samriddha Mondal

Student

★★★★★

"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."

Vansh Agarwal

Student

★★★★★

"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."

Sonam Oswal

Parent of Dhairya

★★★★★

"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."

Sonu Goyal

Parent

★★★★★

"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."

Ritu Kedia

Parent

The rest of the series

Fourteen more pages on vibe coding and learning AI

This page is the project ladder. The definition, the age tracks and the deciding-parent pages each stand on their own.

Questions about student projects

What students and parents ask about project work

Where do good project ideas actually come from?

From irritation, almost always. The best student projects we see solve something that annoys the student weekly: a sibling dispute, a lost permission slip, a team whose fixtures nobody can find. Ideas from lists, including this one, work best as templates you re-aim at your own life. An idea the student did not choose survives about two weeks; an irritation they personally feel survives the debugging.

My child rebuilds things from YouTube. Does that count as a project?

It counts as practice, and practice is fine. It becomes a project the moment it diverges: a feature the video never built, a redesign for a real user, a version that handles the messy cases the tutorial skipped. Our teachers force that divergence deliberately, because the divergence is where following ends and building begins.

How long should a student project take?

First-rung builds are an afternoon. Stretch builds are a weekend or two. A term project deserves about three weeks of evenings, and our 21-day plan on this page exists because that length forces real scoping without exhausting a school-term attention span. Anything projected to take months should be cut down until it fits, not because ambition is bad but because finishing is the skill being trained.

Should student projects go on GitHub?

From the teen years onward, yes, and earlier than most people expect. Version history is the honest diary of a project: it proves the work happened over time, it makes wrong turns recoverable, and it is the difference between a folder of files and something you can show. We teach Git in the teen track for exactly this reason, and there is a dedicated Git and GitHub course when a student is ready to do it properly.

Do competitions and science fairs allow AI-assisted projects?

Policies vary by event and are changing every season, so the only safe rule is the one we teach: disclose the AI use plainly, and be able to explain every part of the project without the AI present. Judges consistently forgive assistance and consistently punish the student who cannot answer what a line of their own project does. Being interrogatable is the real requirement, whatever the written policy says.

Can two friends build one project together?

Yes, with one condition that saves the friendship and the project: split by parts, not by turns. One owns the data and logic, the other owns the interface, and they integrate weekly. Taking turns at one keyboard produces a project neither can explain and an argument about whose idea it was. Split ownership produces two students who each own something completely.

What makes a project stand out to a reviewer?

Four signals, in the order reviewers meet them: it runs, live, at a link that works on their phone; a short written note explains one decision and one thing that went wrong; a real person other than the builder has used it; and the builder can tour the code cold. Notice that none of these is visual polish. A plain tool with those four beats a beautiful demo missing them, in classrooms and interviews alike.

My student has five abandoned projects. Is that normal?

Normal, and worth changing. The pile usually means projects are being chosen four sizes too large, which vibe coding encourages because starting is now so cheap. The fix is not discipline lectures, it is the finishing experience: one deliberately small project taken all the way to shipped changes how a student chooses forever after. That is why our terms end in demo days rather than in syllabus completion.

What do project-based classes cost here?

Group batches of five to eight are USD 100 a month and one to one is USD 150 a month, the same in every country we teach outside India, with no enrolment charge and no minimum stay. Every track is project-based; it is not a premium format we charge extra for, it is just how we think teaching works.

How does a student start with a project already in mind?

Perfectly: send the form and mention the idea. The free first class starts scoping it on the spot, which is the fastest possible demonstration of how we work, and the student leaves with the idea cut to a finishable size and its first slice already running. Arriving with no idea is equally fine; rung one of the ladder above exists precisely for that.

Start

Arrive with an idea, leave with it scoped and running

The free class on this page has a specific shape: bring the project, any project, and watch a teacher run the three scoping questions on it live, then build the first slice with the student before the call ends. It is the 21-day plan's first two days, compressed into ninety minutes, and it works as an honest sample of everything we do. No idea yet? We keep a shelf of rung-one irritations ready, and every student turns out to have one of their own within ten minutes of being asked the right way.

More background reading: what vibe coding actually is, the teen track where the ladder lives, or what our students have built.

WhatsApp us · +91 91233 66161 · contact@modernagecoders.com

Teaching is live video from India, worldwide, and the number above is an Indian line. Sending the form books nothing and costs nothing; it starts the phone call in which everything else gets decided by you.

No card. No obligation. We call, we do not spam.

Keep exploring Modern Age Coders

Coding classes in nearby places

Learn more

Free resources

From the blog

Start here