glass/PLAN.md

Project Plan: Firebase Integration & Encryption

This document outlines the plan to integrate Firebase Firestore as a remote database for logged-in users and implement end-to-end encryption for user data.

Phase 1: encryptionService and Secure Key Management

The goal of this phase is to create a centralized service for data encryption and decryption, with secure management of encryption keys.

1. Install keytar: Add the keytar package to the project to securely store encryption keys in the OS keychain.
2. Create encryptionService.js:
   • Location: src/common/services/encryptionService.js
   • Implement encrypt(text) and decrypt(encrypted) functions using Node.js crypto with AES-256-GCM.
3. Implement Key Management:
   • Create an initializeKey(userId) function within the service.
   • This function will first attempt to retrieve the encryption key from keytar.
   • If keytar fails or no key is found, it will generate a secure, session-only key in memory as a fallback. It will not save the key to an insecure location like electron-store.

Phase 2: Automatic Encryption/Decryption via Firestore Converter

This phase aims to abstract away the encryption/decryption logic from the repository layer, making it automatic.

1. Create firestoreConverter.js:
   • Location: src/common/repositories/firestoreConverter.js
   • Implement a factory function createEncryptedConverter(fieldsToEncrypt = []).
   • This function will return a Firestore converter object with toFirestore and fromFirestore methods.
   • toFirestore: Will automatically encrypt the specified fields before writing data to Firestore.
   • fromFirestore: Will automatically decrypt the specified fields after reading data from Firestore.

Phase 3: Implement Firebase Repositories

With the encryption layer ready, we will create the Firebase equivalents of the existing SQLite repositories.

1. Create session/firebase.repository.js:
   • Location: src/common/repositories/session/firebase.repository.js
   • Use the createEncryptedConverter to encrypt fields like title.
   • Implement all functions from the SQLite counterpart (create, getById, getOrCreateActive, etc.) using Firestore APIs.
2. Create ask/repositories/firebase.repository.js:
   • Location: src/features/ask/repositories/firebase.repository.js
   • Use the createEncryptedConverter to encrypt the content field of AI messages.
   • Implement all functions from the SQLite counterpart (addAiMessage, getAllAiMessagesBySessionId).

Phase 4: Integrate Repository Strategy Pattern

This final phase will activate the logic that switches between local and remote databases based on user authentication status.

1. Update getRepository() functions:
   • Modify src/common/repositories/session/index.js and src/features/ask/repositories/index.js.
   • In the getRepository() function, use authService.getCurrentUser() to check if the user is logged in (user.isLoggedIn).
   • If logged in, return the firebaseRepository.
   • Otherwise, return the sqliteRepository.
   • Uncomment the require statements for the newly created Firebase repositories.