MVP auth methods and Free/Pro/Pro Max entitlement baseline

DEC-0033 - MVP auth methods and Free/Pro/Pro Max entitlement baseline

Superseded by DEC-0044 (auth/identity model refinement) and DEC-0048 (tier/credit visibility refinement) on 2026-02-18.

Context

After confirming single-user and tier taxonomy baseline (DEC-0032), the product still needed concrete signup methods and tier-level capability boundaries for web V2.

Decision

MVP authentication methods:

• Allowed methods:
  • Vietnam phone number via OTP,
  • email via OTP,
  • Google sign-in,
  • Facebook sign-in.
• For phone and email registration, verification is OTP-based.

Tier entitlement baseline:

• Free:
  • can attempt all exercises,
  • can use non-AI platform features,
  • Reading/Listening score visibility is allowed,
  • Writing/Speaking AI grading and AI explanation are not available.
• Pro:
  • unlimited Reading/Listening practice,
  • unlimited Writing/Speaking attempt with basic scoring,
  • Pro Linearthinking explanation for Reading/Listening,
  • access to core special features (Practice Management, learning stats, vocabulary notebook).
• Pro Max:
  • unlimited practice across all 4 skills,
  • includes AI credits for advanced features.

Decision Value

• Locks one clear auth set for MVP implementation and QA.
• Prevents Free/Pro/Pro Max confusion at feature gate points.
• Keeps value ladder simple: non-AI baseline -> Pro core -> Pro Max AI depth.

Rationale

This gives a practical boundary that matches user intent: broad practice access for everyone, paid differentiation on explanation depth and AI capability.