feat: Excel template export with meeting number auto-generation

- Add meeting_number field (M-YYYYMMDD-XX format) with auto-generation
- Refactor Excel export to use cell coordinates instead of placeholders
- Export files saved to backend/record/ directory with meeting number filename
- Add database migration for meeting_number column
- Add start.sh script for managing frontend/backend/sidecar services
- Update OpenSpec documentation

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

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

openspec/changes/archive/2025-12-11-update-excel-template-export/specs/ai-summarization/spec.md

@@ -0,0 +1,31 @@
+## MODIFIED Requirements
+
+### Requirement: Structured Output Format
+The AI summarization SHALL return structured data with conclusions and action items.
+
+#### Scenario: Complete structured response
+- **WHEN** transcript contains clear decisions and assignments
+- **THEN** response SHALL include conclusions array (0 to many items) and action_items array with content, owner, due_date fields
+
+#### Scenario: Partial data extraction
+- **WHEN** transcript lacks explicit owner or due_date for action items
+- **THEN** those fields SHALL be empty strings or null allowing manual completion
+
+#### Scenario: Variable conclusion count
+- **WHEN** transcript has multiple decisions
+- **THEN** conclusions array SHALL contain all extracted conclusions without artificial limit
+
+#### Scenario: No conclusions found
+- **WHEN** transcript has no clear decisions
+- **THEN** conclusions array SHALL be empty []
+
+### Requirement: Dify Prompt Configuration
+The Dify workflow SHALL be configured with detailed system prompt for meeting summarization.
+
+#### Scenario: System prompt behavior
+- **WHEN** transcript is sent to Dify
+- **THEN** Dify SHALL use configured system prompt to extract conclusions and action_items in JSON format
+
+#### Scenario: Flexible output count
+- **WHEN** Dify processes transcript
+- **THEN** it SHALL return variable number of conclusions and action items based on actual content