* Add AI intelligence system with memory, sessions, skills and constitution
Implements a client-side AI system inspired by OpenClaw's architecture:
Rust backend (src-tauri/src/ai/):
- SQLite memory database with FTS5 full-text search (memory_db.rs)
- AES-256-GCM encryption with Argon2id key derivation (encryption.rs)
- Session JSONL file I/O and memory file management (sessions.rs)
- 30 new Tauri commands for memory, encryption, and session operations
TypeScript AI modules (src/lib/ai/):
- Constitution framework: safety rules, memory principles, action validation
- Memory system: markdown chunking, hybrid FTS5+vector search, encryption
- Prompt system: modular builder with 7 sections (constitution, identity,
crypto-intelligence, memory-recall, skills, tools, context)
- Session system: JSONL transcripts, context compaction, memory flush
- Skills framework: SKILL.md loader, registry, GitHub installer
- LLM providers: pluggable interface with custom and OpenAI implementations
- AI tools: memory_search, memory_read, memory_write, web_search
Integration:
- Redux aiSlice with persisted config
- AIProvider in the app provider chain
- All data encrypted at rest, zero-knowledge architecture
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Apply cargo fmt formatting
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Add unit tests for AI intelligence system
Cover constitution (loader, validator), memory (chunker), prompts
(system-prompt, sections), providers (embeddings), sessions (types),
skills (frontmatter, registry), tools (registry), and Redux aiSlice.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
* Enhance AI system with entity management and constitution updates
- Introduced EntityManager and EntityQuery classes for managing entity relationships and queries within the AI platform.
- Updated constitution loader to fetch the constitution from a public GitHub repository, with fallback to a bundled default.
- Enhanced constitution parsing to support new formats and improved error handling.
- Updated memory schema to reflect changes in storage management.
- Added support for loading skill definitions from TypeScript files, improving skill lifecycle management.
This update strengthens the AI system's capabilities in entity management and constitution handling, ensuring better compliance and functionality.
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Implements a sync service that loads all chats via GetDialogs in batched
100-chat requests, preloads messages for the top 20 chats with rate-limit
pacing, and processes real-time Telegram updates through the existing
UpdateManager's PTS/SEQ state machine. The service is triggered automatically
by TelegramProvider when the connection is established and authenticated.
New files: syncService, entityBuilders, updateHandler, schedulers (all under
src/services/telegram/). Redux state extended with isSyncing, isSynced, users
map, and batch chat reducers (replaceChats, addChats, setSyncStatus, etc.).
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Changed all references of the deep link scheme in documentation and code to `alphahuman://` to align with the new branding.
- Updated the Telegram login flow documentation to reflect the new deep link handling.
- Adjusted various files to ensure consistency in the authentication process across platforms.
This update enhances the clarity and branding of the authentication flow for the desktop application.
- Introduced detailed architecture overview outlining system components and entry points.
- Documented state management using Redux Toolkit and Redux-Persist, including store configuration and slice structures.
- Described services layer architecture, detailing API client, socket service, and MTProto service functionalities.
- Explained the Model Context Protocol (MCP) system, including tool categories and implementation examples.
- Outlined routing structure and page components, emphasizing protected and public routes.
- Organized reusable components by feature, detailing their structure and usage.
- Provided insights into provider management for service lifecycle and shared state.
This documentation enhances understanding of the application's structure and improves onboarding for new developers, ensuring clarity in the system's design and functionality.
