feat: Improve file display, timezone handling, and LOT management

Changes:
- Fix datetime serialization with UTC 'Z' suffix for correct timezone display
- Add PDF upload support with extension fallback for MIME detection
- Fix LOT add/remove by creating new list for SQLAlchemy JSON change detection
- Add file message components (FileMessage, ImageLightbox, UploadPreview)
- Add multi-file upload support with progress tracking
- Link uploaded files to chat messages via message_id
- Include file attachments in AI report generation
- Update specs for file-storage, realtime-messaging, and ai-report-generation

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

openspec/specs/ai-report-generation/spec.md

@@ -50,7 +50,7 @@ The system SHALL maintain a permanent `users` table to store user display names
 
 ### Requirement: Report Data Collection
 
-The system SHALL collect all relevant room data for AI processing, including messages, members, files, and room metadata.
+The system SHALL collect all relevant room data for AI processing, including messages, members, files with their conversation context, and room metadata.
 
 #### Scenario: Collect complete room data for report generation
 
@@ -62,9 +62,29 @@ The system SHALL collect all relevant room data for AI processing, including mes
   - Room metadata (title, incident_type, severity, status, location, description, timestamps)
   - All 50 messages sorted by created_at ascending
   - All 5 members with their roles (owner, editor, viewer)
-  - All 3 files with metadata (filename, type, uploader, upload time)
+  - All 3 files with metadata (filename, type, uploader, upload time) AND their associated message context
 - **AND** messages SHALL include sender display name (not just user_id)
-- **AND** file references in messages SHALL be annotated as "[附件: filename.ext]"
+- **AND** file references in messages SHALL be annotated with surrounding context
+
+#### Scenario: Include file context in report data
+- **GIVEN** a file "defect_photo.jpg" was uploaded with the message "發現產品表面瑕疵"
+- **AND** the previous message was "Line 3 溫度異常升高中"
+- **AND** the next message was "已通知維修人員處理"
+- **WHEN** report data is collected
+- **THEN** the file entry SHALL include:
+  ```json
+  {
+    "file_id": "...",
+    "filename": "defect_photo.jpg",
+    "uploader_display_name": "陳工程師",
+    "uploaded_at": "2025-12-08T14:30:00+08:00",
+    "caption": "發現產品表面瑕疵",
+    "context_before": "Line 3 溫度異常升高中",
+    "context_after": "已通知維修人員處理"
+  }
+  ```
+- **AND** the AI prompt SHALL format files as:
+  `[附件: defect_photo.jpg] - 上傳者: 陳工程師 (14:30), 說明: "發現產品表面瑕疵" (前文: "Line 3 溫度異常升高中")`
 
 #### Scenario: Handle room with no messages
 
@@ -82,8 +102,6 @@ The system SHALL collect all relevant room data for AI processing, including mes
 - **AND** summarize older messages by day (e.g., "2025-12-01: 45 則訊息討論設備檢修")
 - **AND** the total formatted content SHALL stay within token limits
 
----
-
 ### Requirement: DIFY AI Integration
 
 The system SHALL integrate with DIFY Chat API to generate structured report content from collected room data.
@@ -126,7 +144,7 @@ The system SHALL integrate with DIFY Chat API to generate structured report cont
 
 ### Requirement: Document Assembly
 
-The system SHALL assemble professional .docx documents from AI-generated content with embedded images from MinIO.
+The system SHALL assemble professional .docx documents from AI-generated content with embedded images from MinIO and file context from conversations.
 
 #### Scenario: Generate complete report document
 
@@ -142,8 +160,9 @@ The system SHALL assemble professional .docx documents from AI-generated content
   - Section 4: 處理過程 (from AI resolution_process.content)
   - Section 5: 目前狀態 (from AI current_status)
   - Section 6: 最終處置結果 (from AI final_resolution, if has_resolution=true)
-  - Section 7: 附件 (embedded images + file list)
+  - Section 7: 附件 (embedded images with captions + file list with context)
 - **AND** images SHALL be embedded at appropriate size (max width 15cm)
+- **AND** each image SHALL include its caption from the upload message
 - **AND** document SHALL use professional formatting (標楷體 or similar)
 
 #### Scenario: Handle missing images during assembly
@@ -163,8 +182,6 @@ The system SHALL assemble professional .docx documents from AI-generated content
 - **THEN** the system SHALL create a complete document without the embedded images section
 - **AND** the attachments section SHALL show "本事件無附件檔案" if no files exist
 
----
-
 ### Requirement: Report Generation API
 
 The system SHALL provide REST API endpoints for triggering report generation and downloading generated reports.