feat: automate .env creation during setup for zero-config onboarding

style: tidy project root and simplify onboarding commands

chore: move src-tauri to src/backend for better logical structure

chore: remove electron-related source files and dependencies

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.

refactor: modernize source code with improved structure and logging ...

- Add config.js for centralized configuration management
- Add logger.js for structured logging with levels
- Modernize main.js with JSDoc comments and error handling
- Enhance mcp-server.js with health check endpoint and logging
- Refactor db.js with validation and error handling functions
- Add TOOL_NAMES constant and improve mcp-tools.js documentation
- Refactor mcp-handlers.js with modular handler functions
- Enhance llama-client.js with timeout and input validation
- Update test to match new error messages
- All 35 tests passing with improved code quality

Add comprehensive test suite (35 tests) and update journal ...

- Create test/ folder with 5 test modules (35 tests total)
  * db.test.js: Database operations (5 tests)
  * mcp-tools.test.js: Tool definitions (11 tests)
  * mcp-handlers.test.js: Handler implementation (6 tests)
  * llama-client.test.js: LLama API integration (9 tests)
  * integration.test.js: End-to-end workflows (4 tests)
- Add test/setup.js: Shared test utilities
- Add test/README.md: Test documentation
- Update package.json: Add test scripts (7 variants)
- Update journals/20260206-0000-案件.md: Document test implementation
- All 35 tests passing (139 expect calls)

Add MCP server with llama.cpp integration and docs