feat: Add Expert Transformation Agent with multi-expert perspective system

- Backend: Add expert transformation router with 3-step SSE pipeline
  - Step 0: Generate diverse expert team (random domains)
  - Step 1: Each expert generates keywords for attributes
  - Step 2: Batch generate descriptions for expert keywords
- Backend: Add simplified prompts for reliable JSON output
- Frontend: Add TransformationPanel with React Flow visualization
- Frontend: Add TransformationInputPanel for expert configuration
  - Expert count (2-8), keywords per expert (1-3)
  - Custom expert domains support
- Frontend: Add expert keyword nodes with expert badges
- Frontend: Improve description card layout (wider cards, more spacing)
- Frontend: Add fallback for missing descriptions with visual indicators

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

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

backend/app/routers/transformation.py

@@ -0,0 +1,116 @@
+"""Transformation Agent 路由模組"""
+
+import json
+import logging
+from typing import AsyncGenerator, List
+
+from fastapi import APIRouter
+from fastapi.responses import StreamingResponse
+
+from ..models.schemas import (
+    TransformationRequest,
+    TransformationCategoryResult,
+    TransformationDescription,
+)
+from ..prompts.transformation_prompt import (
+    get_keyword_generation_prompt,
+    get_batch_description_prompt,
+)
+from ..services.llm_service import ollama_provider, extract_json_from_response
+
+logger = logging.getLogger(__name__)
+router = APIRouter(prefix="/api/transformation", tags=["transformation"])
+
+
+async def generate_transformation_events(
+    request: TransformationRequest
+) -> AsyncGenerator[str, None]:
+    """Generate SSE events for transformation process"""
+    try:
+        temperature = request.temperature if request.temperature is not None else 0.7
+        model = request.model
+
+        # ========== Step 1: Generate new keywords ==========
+        yield f"event: keyword_start\ndata: {json.dumps({'message': f'為「{request.category}」生成新關鍵字...'}, ensure_ascii=False)}\n\n"
+
+        keyword_prompt = get_keyword_generation_prompt(
+            category=request.category,
+            attributes=request.attributes,
+            keyword_count=request.keyword_count
+        )
+        logger.info(f"Keyword prompt: {keyword_prompt[:200]}")
+
+        keyword_response = await ollama_provider.generate(
+            keyword_prompt, model=model, temperature=temperature
+        )
+        logger.info(f"Keyword response: {keyword_response[:500]}")
+
+        keyword_data = extract_json_from_response(keyword_response)
+        new_keywords = keyword_data.get("keywords", [])
+
+        yield f"event: keyword_complete\ndata: {json.dumps({'keywords': new_keywords}, ensure_ascii=False)}\n\n"
+
+        if not new_keywords:
+            yield f"event: error\ndata: {json.dumps({'error': '無法生成新關鍵字'}, ensure_ascii=False)}\n\n"
+            return
+
+        # ========== Step 2: Generate descriptions for each keyword ==========
+        yield f"event: description_start\ndata: {json.dumps({'message': '生成創新應用描述...'}, ensure_ascii=False)}\n\n"
+
+        # Use batch description prompt for efficiency
+        desc_prompt = get_batch_description_prompt(
+            query=request.query,
+            category=request.category,
+            keywords=new_keywords
+        )
+        logger.info(f"Description prompt: {desc_prompt[:300]}")
+
+        desc_response = await ollama_provider.generate(
+            desc_prompt, model=model, temperature=temperature
+        )
+        logger.info(f"Description response: {desc_response[:500]}")
+
+        desc_data = extract_json_from_response(desc_response)
+        descriptions_raw = desc_data.get("descriptions", [])
+
+        # Convert to TransformationDescription objects
+        descriptions: List[TransformationDescription] = []
+        for desc in descriptions_raw:
+            if isinstance(desc, dict) and "keyword" in desc and "description" in desc:
+                descriptions.append(TransformationDescription(
+                    keyword=desc["keyword"],
+                    description=desc["description"]
+                ))
+
+        yield f"event: description_complete\ndata: {json.dumps({'count': len(descriptions)}, ensure_ascii=False)}\n\n"
+
+        # ========== Build final result ==========
+        result = TransformationCategoryResult(
+            category=request.category,
+            original_attributes=request.attributes,
+            new_keywords=new_keywords,
+            descriptions=descriptions
+        )
+
+        final_data = {
+            "result": result.model_dump()
+        }
+        yield f"event: done\ndata: {json.dumps(final_data, ensure_ascii=False)}\n\n"
+
+    except Exception as e:
+        logger.error(f"Transformation error: {e}")
+        yield f"event: error\ndata: {json.dumps({'error': str(e)}, ensure_ascii=False)}\n\n"
+
+
+@router.post("/category")
+async def transform_category(request: TransformationRequest):
+    """處理單一類別的轉換"""
+    return StreamingResponse(
+        generate_transformation_events(request),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )