Skip to content

Add Cartoon Vocab MVP product requirements document - #1

Open
ericake wants to merge 5 commits into
masterfrom
claude/admiring-bohr-ga0l7j
Open

ericake wants to merge 5 commits into
masterfrom
claude/admiring-bohr-ga0l7j

Conversation

@ericake

@ericake ericake commented Oct 12, 2026

Copy link
Copy Markdown
Owner

Summary

Add comprehensive product requirements document (PRD) for the Cartoon Vocab MVP, a web application designed to help 6-year-old children learn English vocabulary from animated shows through pre-viewing and post-viewing lessons using sound-to-image mapping.

Key Changes

  • Document Structure: Complete PRD covering background, goals, user roles, core workflows, and detailed functional requirements

  • Core Features Defined:

    • Parent dashboard for episode management, transcript input, vocabulary extraction, and learning report generation
    • Child-friendly interface with zero-text design, English-only voice guidance, and sound-to-image recognition exercises
    • Automatic vocabulary extraction from transcripts with AI-powered selection and human review workflow
    • Multi-stage learning modes: introduction, 2-choice recognition, 4-choice recognition, sentence context, and spaced repetition
    • Gamification through sticker rewards and streak tracking
  • Content Generation Pipeline: Detailed specifications for transcript cleaning, NLP-based vocabulary extraction, LLM-powered selection, image generation, and TTS audio production

  • Success Metrics: Defined KPIs for efficiency (≤3 min parent prep), learning outcomes (≥80% recognition accuracy), and user engagement (≥4 days/week usage, 50% 4-week retention)

  • MVP Scope: Clearly delineated P0 (must-have), P1 (nice-to-have), and P2 (future) features; first launch content set to Peppa Pig Season 1 Episode 1 "Muddy Puddles"

  • Design Principles: Established interaction guidelines emphasizing zero-text interface, English-only voice, large touch targets, no punishment feedback, and consistent task flow

Notable Implementation Details

  • Child Safety: Compliance with COPPA and Chinese child data protection regulations; no real names, photos, or voice recording; no ads or external links
  • Offline Support: Generated courses cached locally for offline completion with sync on reconnection
  • Data Model: Defined entities for child profiles, episodes, vocabulary items, learning sessions, and mastery tracking with spaced repetition scheduling
  • Risk Mitigation: Addressed AI image accuracy, vocabulary visualization challenges, copyright concerns, and child engagement safeguards
  • Confirmed Decisions: Documented key product decisions including English-only voice guidance, non-linear preview/review flow, 10-item default per episode, and web app (PWA) platform choice

https://claude.ai/code/session_01Hih1UmpuHqBoPB9VDZpkhK

English-only voice guidance, no forced preview/review order, 10 items
per episode by default, Peppa Pig S1E1 as launch content, web app,
no per-episode generation cost cap.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Hih1UmpuHqBoPB9VDZpkhK
Covers architecture, data model, REST API, lesson package format,
content pipeline (Claude extraction, ByteDance seed-audio TTS, image
provider abstraction), frontend player/offline design, security and
test plan.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Hih1UmpuHqBoPB9VDZpkhK
FastAPI backend: transcript cleaning, rule-based + Claude word selection,
ByteDance seed-audio TTS client, placeholder image provider, lesson
builder, spaced review scheduler, reports. React PWA frontend with
zero-text child player and parent console.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Hih1UmpuHqBoPB9VDZpkhK
…cript

- Lesson builder never pairs items that share a content word
  (puddle / muddy puddle, jump / jump up and down, mud / muddy)
- backend/.env support and .env.example
- scripts/gen_system_voice.py to pre-generate English prompt audio
- cartoon-vocab/README.md with run instructions and v1 deviations

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Hih1UmpuHqBoPB9VDZpkhK
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants