feat: 日本語 LSA セマンティック検索の実装 (Lindera + SVD) ...

- Lindera による日本語形態素解析トークナイザーを追加
- ndarray-linalg / rsvd による LSA (潜在意味解析) コアロジックを実装
- SQLite テーブル (items_lsa) へのベクトル保存機能を追加
- MCP ツール (lsa_search, lsa_retrain) を統合
- 「軽さは正義」に基づき BM25 (FTS5) 依存を排除し、LSA に一本化

feat: implement resident mode and mcp activity visualization for v0.2.5 release

feat: 文章の登録・更新・削除ツールの実装 (llama-serverエンベディング連携)

サイドカー自動起動・監視、llama_server状態API、UI連携、ESLint/Prettier整備、journal追加

chore: comprehensive project restructuring and cleanup ...

- Moved backend source to src/backend
- Consolidated build artifacts to root target/
- Organized assets to resources/ and documentation to document/
- Updated README with detailed architecture and simplified onboarding
- Fixed minor bug in lib.rs database initialization
- Cleaned up obsolete Electron-era files and test scripts

fix: align all embedding dimensions to 384 for Gemma 3 300M consistency

feat: add local datetime support and automated updated_at trigger for items table

chore: migrate to Tauri/Rust backend with SeaORM and add unit/integration tests ...

- Migrated Electron/Bun backend to Rust (Tauri v2).
- Integrated SeaORM with SQLite and sqlite-vec support.
- Implemented unit tests for LlamaClient and database initialization.
- Added integration tests for MCP handlers.
- Updated documentation (README.md, walkthrough, implementation plan).
- Improved dev environment with launch scripts and tasks.
- Followed user global rules for journaling and gitignore.