From 223227320008857beed7fd5c63dda25af7169be4 Mon Sep 17 00:00:00 2001
From: jaehwang <jhyeu@o2o.kr>
Date: Tue, 2 Jun 2026 16:21:19 +0900
Subject: [PATCH] =?UTF-8?q?transformation=20=EC=84=B9=EC=85=98=20=EC=A0=95?=
 =?UTF-8?q?=EB=B3=B4=20=EC=B6=9C=EB=A0=A5?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 app/integrations/llm/prompt.py                 |  8 ++++++++
 app/integrations/llm/schemas/report.py         |  7 +++++++
 .../llm/temp-prompt/transformation_prompt.txt  | 14 ++++++++++++++
 app/services/analysis.py                       | 18 ++++++++++++++++--
 4 files changed, 45 insertions(+), 2 deletions(-)
 create mode 100644 app/integrations/llm/temp-prompt/transformation_prompt.txt

diff --git a/app/integrations/llm/prompt.py b/app/integrations/llm/prompt.py
index 3ee76d4..4d6a06b 100644
--- a/app/integrations/llm/prompt.py
+++ b/app/integrations/llm/prompt.py
@@ -6,6 +6,7 @@ from integrations.llm.schemas.report import (
     CriticalIssuesInput, CriticalIssuesOutput,
     YouTubeDiagnosisInput, YouTubeDiagnosisOutput,
     BrandConsistencyInput, BrandConsistencyOutput,
+    TransformationInput, TransformationProposal,
 )
 from integrations.llm.schemas.plan import PlanInput, PlanOutput
 from integrations.llm.schemas.market import (
@@ -106,3 +107,10 @@ critical_issues_prompt = Prompt(
     input_class=CriticalIssuesInput,
     output_class=CriticalIssuesOutput,
 )
+
+transformation_prompt = Prompt(
+    file_name="transformation_prompt.txt",
+    prompt_model="REPORT_MODEL",
+    input_class=TransformationInput,
+    output_class=TransformationProposal,
+)
diff --git a/app/integrations/llm/schemas/report.py b/app/integrations/llm/schemas/report.py
index 8bbee91..f9dc896 100644
--- a/app/integrations/llm/schemas/report.py
+++ b/app/integrations/llm/schemas/report.py
@@ -368,6 +368,13 @@ class CriticalIssuesOutput(BaseModel):
     diagnosis: list[DiagnosisItem]
 
 
+# --- Transformation ---
+
+class TransformationInput(BaseModel):
+    clinic_name: str | None = None
+    data: str | None = None
+
+
 # --- BrandConsistency ---
 
 class BrandConsistencyInput(BaseModel):
diff --git a/app/integrations/llm/temp-prompt/transformation_prompt.txt b/app/integrations/llm/temp-prompt/transformation_prompt.txt
new file mode 100644
index 0000000..58beb39
--- /dev/null
+++ b/app/integrations/llm/temp-prompt/transformation_prompt.txt
@@ -0,0 +1,14 @@
+다음은 성형외과/피부과 {clinic_name} 의 전 채널 수집 데이터입니다.
+
+{data}
+
+위 데이터를 바탕으로 이 병원의 마케팅 전환 전략을 수립해줘.
+아래 5개 항목을 포함한 JSON을 출력해줘.
+
+1. brand_identity: 브랜드 아이덴티티 개선 항목 (area, as_is, to_be)
+2. content_strategy: 콘텐츠 전략 개선 항목 (area, as_is, to_be)
+3. platform_strategies: 플랫폼별 전략 (platform, icon, current_metric, target_metric, strategies[{strategy, detail}])
+4. website_improvements: 웹사이트 개선 항목 (area, as_is, to_be)
+5. new_channel_proposals: 신규 채널 제안 (channel, priority, rationale)
+
+출처 번호([1], [2] 등)는 포함하지 마.
diff --git a/app/services/analysis.py b/app/services/analysis.py
index 48dcea9..b6d67cc 100644
--- a/app/services/analysis.py
+++ b/app/services/analysis.py
@@ -8,8 +8,8 @@ from common.db.run import select_run, update_run_report, update_run_plan
 from common.db.source import select_run_raw_data, select_run_mainpage_url
 from common.db.market import select_market
 from integrations.llm.llm_service import LLMService
-from integrations.llm.prompt import report_prompt, plan_prompt, youtube_diagnosis_prompt, brand_consistency_prompt, critical_issues_prompt
-from integrations.llm.schemas.report import ReportOutput, ClinicSnapshot, YouTubeAudit, BrandConsistencyOutput, CriticalIssuesOutput, DiagnosisItem
+from integrations.llm.prompt import report_prompt, plan_prompt, youtube_diagnosis_prompt, brand_consistency_prompt, critical_issues_prompt, transformation_prompt
+from integrations.llm.schemas.report import ReportOutput, ClinicSnapshot, YouTubeAudit, BrandConsistencyOutput, CriticalIssuesOutput, DiagnosisItem, TransformationProposal
 from integrations.llm.schemas.plan import PlanOutput
 
 logger = logging.getLogger(__name__)
@@ -208,6 +208,17 @@ async def _build_youtube_audit(youtube: dict) -> dict:
     return YouTubeAudit.model_validate(yt_patch).model_dump()
 
 
+async def _build_transformation(analysis_run_id: str, raw: dict) -> dict:
+    result: TransformationProposal = await LLMService(provider="perplexity").generate(
+        transformation_prompt,
+        {
+            "clinic_name": (raw.get("mainpage") or {}).get("clinicName"),
+            "data":        json.dumps(raw, ensure_ascii=False),
+        },
+    )
+    return result.model_dump()
+
+
 async def _build_critical_issues(analysis_run_id: str, raw: dict) -> list[dict]:
     result: CriticalIssuesOutput = await LLMService(provider="perplexity").generate(
         critical_issues_prompt,
@@ -255,6 +266,7 @@ async def _build_overrides(analysis_run_id: str) -> dict:
     brand = await generate_brand_consistency(analysis_run_id)
     brand_patch = brand.model_dump()["brand_inconsistencies"]
     critical_issues = await _build_critical_issues(analysis_run_id, raw)
+    transformation = await _build_transformation(analysis_run_id, raw)
 
     fb_patch: dict = {}
     if fb_pages:
@@ -273,6 +285,8 @@ async def _build_overrides(analysis_run_id: str) -> dict:
         overrides["youtube_audit"] = yt_patch
     if critical_issues:
         overrides["problem_diagnosis"] = critical_issues
+    if transformation:
+        overrides["transformation"] = transformation
     return overrides