Mobile: Companion app — quick capture + search

[ajianaz]

Description

Mobile companion for quick memory capture and search on-the-go.

Features

• Quick capture: text, voice, photo -> memory
• Semantic search over synced knowledge base
• Offline support (local cache)

Architecture Decision

• Framework: Flutter (Impeller, Dart) - see analysis below
• Repo: codecoradev/corin-mobile (org-owned)
• Backend: Corin Web API (GH#21 prerequisite)
• Offline: Drift (SQLite ORM) + Dio sync queue

Why Flutter (not Tauri Mobile)

Criterion	Tauri 2 Mobile	Flutter
Sibung expertise	No	Yes - Flutter dev
Offline-first ecosystem	Primitive (localStorage)	Mature (Drift, Isar, Hive)
Camera/Voice plugins	Experimental	Production-ready
Rendering	WebView (limited perf)	Impeller (native 60fps)
Mobile ecosystem maturity	New (Oct 2024)	8+ years, massive

Tauri mobile code-sharing with desktop is a trap -- mobile UI needs fundamentally different patterns (swipe, FAB, bottom sheets). Real code sharing happens at the backend (Uteke HTTP API), not the frontend.

Prerequisites

BLOCKED by GH#21. Mobile cannot start until Corin Web API ships.

• GH#21: Corin Web API for remote/mobile access
• Corin v0.3.0+ with stable Web API endpoints
• Web API spec finalized and tested

Tech Stack

Layer	Choice
Framework	Flutter 3.27+ (Impeller)
State	Cubit
Local DB	Drift
Network	Dio + retrofit
Models	freezed
Monorepo	Melos
Minimum	iOS 15+, Android API 24+

Architecture

Mobile (Flutter)
  Capture | Search | Settings
  Drift (local) | Dio (Corin Web API GH#21)
       |               |
  Local SQLite     Corin Desktop (LAN)
  (offline cache)

Implementation Roadmap

1. Phase A (Prerequisite): GH#21 - Corin Web API (~2-3 weeks)
2. Phase B (Foundation): Flutter project setup, server config, Drift schema (~2 days)
3. Phase C (Core): Text/voice capture, search, memory list, offline queue (~5 days)
4. Phase D (Polish): Settings, error handling, theme, testing (~5 days)

Total to mobile MVP: ~5-7 weeks from now (assuming Web API ships in 2-3 weeks).

[ajianaz]

Architecture Analysis (2026-07-06)

Full analysis document produced: 15KB covering tech decision, architecture, feature scope, prerequisites, and implementation roadmap.

Key Decisions

1. Flutter (not Tauri Mobile) -- Sibung expertise + mature offline ecosystem + proven native API plugins
2. Repo: codecoradev/corin-mobile (org-owned, not ajianaz/hub-mobile)
3. BLOCKED by GH#21 (Corin Web API) -- no mobile work until Web API ships
4. 5-7 weeks total to working mobile MVP from today

Critical Path

GH#21 (Web API, 2-3 weeks) -> corin-mobile repo creation -> Flutter implementation (3-4 weeks)

Why Not Tauri Mobile

Code-sharing with desktop is illusory for this use case. Mobile needs fundamentally different UI patterns (swipe, FAB, bottom sheets, touch-optimized). Tauri mobile WebView performance and plugin ecosystem are still immature. Real code sharing is at the backend layer (Uteke HTTP API) which both apps already share.