A child-first, scenario-based digital safety simulator that trains scam detection as a daily habit, not a one-time lecture.
Children are entering digital spaces earlier, while scam tactics are becoming faster, more persuasive, and more emotionally manipulative.
Traditional safety education is often passive (slides, warnings, one-way talks). SmartPath Kids turns this into active practice:
- learners investigate realistic content,
- make judgment calls under uncertainty,
- get immediate feedback,
- and build confidence through repetition.
This project is designed to answer one central question:
How can we make critical thinking feel as engaging as the apps children already use?
SmartPath Kids is a gamified web app for primary/secondary learners to practice spotting:
- phishing emails,
- suspicious social posts,
- misleading video descriptions,
- fake urgency/persuasion tactics,
- and “too good to be true” offers.
Instead of telling students what not to do, we let them investigate, decide, and reflect.
-
Simulation over instruction
The UI mirrors familiar digital contexts (News, Social, YouTube, Email), so skill transfer to real life is stronger. -
Micro-investigation mechanics
Each scenario is split into “investigation zones” where users inspect clues and classify as legit or scam. -
Learning economy with consequences
Correct decisions reward coins; incorrect decisions carry penalty (-5) to reinforce careful thinking over guessing. -
Safe AI-assisted variety
Optional AI-generated scenarios create fresh mission content while preserving age-appropriate constraints and safety framing. -
Meta-cognitive reflection
Investigation assistant prompts students to explain why they think something is safe/scam, not just click answers.
- Open landing page and enter app shell.
- Choose a category (
News,Social,Email,YouTube,Random). - Investigate highlighted clues and answer Legit vs Scam/Fake.
- Receive instant feedback and coin updates.
- Use
Game,Ranking, andProfileloops for replayability and progression.
- News: article-style misinformation and source credibility checks.
- Social: post/engagement/link-level manipulation cues.
- YouTube: title/channel/description trust signals.
- Email: sender, subject, urgency, and request legitimacy checks.
- Random Mission: mixed scenario training across categories.
- Coins + scoring
- Correct judgment:
+10 - Incorrect judgment:
-5
- Correct judgment:
- Daily reward
- Achievements
- Ranking by coins and accuracy
- Game mode with fair vs scam offers and mini challenge loops
- AI scenario generation for content freshness
- Investigation assistant for guided questioning
- Phrase-avoidance + variety themes to reduce repetitive content
- React 18
- TypeScript
- Vite
- React Router v6
- CSS Modules
- localStorage-backed state persistence
src/
components/
investigate/ # Investigate zone blocks + assistant chat UI
layout/ # Shell and app grid
context/
AppState.tsx # Coins, accuracy, achievements, daily rewards
data/
mockContent.ts # Base scenarios
gameData.ts # Game offers/challenges
pages/
Index.tsx # Landing page
AppPage.tsx # App shell/home
News/Social/Email/Youtube/Random/Ranking/Profile/Game
services/
aiClient.ts # AI scenario + assistant integration
questionBank.ts # Scenario storage/variety handling
| Route | Purpose |
|---|---|
/ |
Landing page |
/app |
App shell |
/news |
News investigation |
/social |
Social investigation |
/email |
Email investigation |
/youtube |
YouTube investigation |
/random |
Random mission |
/ranking |
Leaderboards |
/profile |
Learner profile |
/game |
Game mode |
- Node.js 18+
- npm
npm install
npm run devOpen: http://localhost:5173
npm run build
npm run previewIf you want AI-generated missions and assistant replies, create .env:
VITE_MINIMAX_API_KEY=your_key_here
VITE_MINIMAX_MODEL=M2-herWithout API keys, base app content still runs using built-in scenarios.
- Age-appropriate language
- No real financial transactions
- No collection of sensitive personal learner data
- Feedback emphasizes reasoning, not fear
- Scam literacy framed as empowerment
SmartPath Kids is not just a “quiz app.”
It is a practical, replayable training environment where children repeatedly exercise judgment in realistic digital contexts. The design combines:
- educational psychology (active retrieval + feedback),
- product design (engaging loops + progression),
- and technical scalability (modular routes + AI-extendable scenario engine).
This makes it both a compelling demo today and a strong foundation for school/community deployment tomorrow.
This project is licensed under the MIT License. See LICENSE.


