feat: GEO data model and API endpoints

- SearchRanking: +5 fields (ai_search_engine, citation_snippet, citation_url, geo_score, content_type)
- New GeoReadinessScore model (6-dimension scoring: schema/faq/howto/citations/word_count/readability/headings)
- GET /api/seo/geo/overview, GET /api/seo/geo/readiness, POST /api/seo/geo/track

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

platform/backend/app/api/search_rankings.py

@@ -5,7 +5,7 @@ from typing import Optional
 from datetime import datetime, timedelta, timezone
 
 from ..database import get_db
-from ..models import SearchRanking, Article, Topic
+from ..models import SearchRanking, Article, Topic, GeoReadinessScore
 from .auth import get_current_user, org_filter
 
 router = APIRouter(prefix="/api/seo", tags=["seo"])
@@ -75,3 +75,84 @@ def get_rankings_overview(
             "best_keyword": best.keyword if best else None,
         }
     }
+
+
+@router.get("/geo/overview")
+def get_geo_overview(
+    days: int = Query(30, ge=1, le=365),
+    db: Session = Depends(get_db),
+    current_user: dict = Depends(get_current_user),
+):
+    """GEO 概览 — AI 搜索引用 + 就绪度评分"""
+    cutoff = datetime.now(timezone.utc) - timedelta(days=days)
+
+    base = db.query(SearchRanking).filter(SearchRanking.checked_at >= cutoff)
+    total_geo = base.filter(SearchRanking.search_engine == "geo").count()
+    total_cited = base.filter(
+        SearchRanking.search_engine == "geo",
+        SearchRanking.ai_cited == True
+    ).count()
+
+    engine_breakdown = {}
+    engines = base.filter(
+        SearchRanking.search_engine == "geo",
+        SearchRanking.ai_source.isnot(None)
+    ).with_entities(
+        SearchRanking.ai_source,
+        func.count(SearchRanking.id).label("cnt"),
+        func.sum(func.cast(SearchRanking.ai_cited, func.Integer)).label("cited"),
+    ).group_by(SearchRanking.ai_source).all()
+    for row in engines:
+        engine_breakdown[row.ai_source] = {
+            "total": row.cnt,
+            "cited": row.cited or 0,
+        }
+
+    latest_scores = db.query(GeoReadinessScore).order_by(
+        desc(GeoReadinessScore.checked_at)
+    ).limit(50).all()
+
+    avg_score = db.query(func.avg(GeoReadinessScore.total_score)).filter(
+        GeoReadinessScore.checked_at >= cutoff
+    ).scalar() or 0
+
+    return {
+        "ok": True,
+        "data": {
+            "total_geo_checks": total_geo,
+            "total_cited": total_cited,
+            "citation_rate": round(total_cited / total_geo * 100, 1) if total_geo else 0,
+            "engine_breakdown": engine_breakdown,
+            "avg_geo_readiness": round(float(avg_score), 1),
+            "latest_scores": [s.to_dict() for s in latest_scores[:10]],
+        }
+    }
+
+
+@router.get("/geo/readiness")
+def get_geo_readiness_scores(
+    days: int = Query(30, ge=1, le=365),
+    limit: int = Query(50, ge=1, le=200),
+    db: Session = Depends(get_db),
+    current_user: dict = Depends(get_current_user),
+):
+    """获取 GEO 就绪度评分列表"""
+    cutoff = datetime.now(timezone.utc) - timedelta(days=days)
+    scores = db.query(GeoReadinessScore).filter(
+        GeoReadinessScore.checked_at >= cutoff
+    ).order_by(desc(GeoReadinessScore.checked_at)).limit(limit).all()
+    return {"ok": True, "data": [s.to_dict() for s in scores]}
+
+
+@router.post("/geo/track")
+def trigger_geo_tracking(
+    db: Session = Depends(get_db),
+    current_user: dict = Depends(get_current_user),
+):
+    """手动触发 GEO 追踪"""
+    try:
+        from scripts.geo_tracker import run_all
+        result = run_all()
+        return {"ok": True, "message": "GEO 追踪已启动", "result": result}
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))

platform/backend/app/models.py

@@ -860,7 +860,7 @@ class CollectorSource(Base):
 
 
 class SearchRanking(Base):
-    """搜索排名追踪"""
+    """搜索排名追踪 + AI 搜索引用追踪"""
     __tablename__ = "search_rankings"
 
     id = Column(Integer, primary_key=True, index=True, autoincrement=True)
@@ -872,7 +872,12 @@ class SearchRanking(Base):
     position = Column(Integer, nullable=True)  # 搜索排名位置（null = 未上榜）
     url_found = Column(String, nullable=True)  # 被找到的具体 URL
     ai_cited = Column(Boolean, default=False)  # 是否被 AI 搜索引用
-    ai_source = Column(String, nullable=True)  # AI 搜索来源名称
+    ai_source = Column(String, nullable=True)  # AI 搜索来源（chatgpt/perplexity/deepseek）
+    ai_search_engine = Column(String, nullable=True)  # AI 搜索具体引擎模型名
+    citation_snippet = Column(String, nullable=True)  # 被引用的文本片段
+    citation_url = Column(String, nullable=True)  # 引用来源链接
+    geo_score = Column(Integer, nullable=True)  # GEO 就绪度评分 0-100
+    content_type = Column(String, nullable=True)  # 内容类型: article/listicle/howto/faq/review
     checked_at = Column(DateTime(timezone=True), server_default=func.now())
 
     def to_dict(self):
@@ -887,5 +892,48 @@ class SearchRanking(Base):
             "url_found": self.url_found,
             "ai_cited": self.ai_cited,
             "ai_source": self.ai_source,
+            "ai_search_engine": self.ai_search_engine,
+            "citation_snippet": self.citation_snippet,
+            "citation_url": self.citation_url,
+            "geo_score": self.geo_score,
+            "content_type": self.content_type,
+            "checked_at": self.checked_at.isoformat() if self.checked_at else None,
+        }
+
+
+class GeoReadinessScore(Base):
+    """GEO 就绪度评分 — 按文章评估被 AI 搜索引用的概率"""
+    __tablename__ = "geo_readiness_scores"
+
+    id = Column(Integer, primary_key=True, index=True, autoincrement=True)
+    article_id = Column(String, ForeignKey("articles.id"), nullable=True, index=True)
+    topic_id = Column(String, ForeignKey("topics.id"), nullable=True)
+    platform = Column(String, nullable=True)
+    total_score = Column(Integer, default=0)  # 总分 0-100
+    has_schema = Column(Boolean, default=False)  # 是否有结构化数据
+    schema_types = Column(String, nullable=True)  # 含有的 schema 类型列表
+    has_faq_format = Column(Boolean, default=False)  # 是否含 FAQ 格式
+    has_howto_format = Column(Boolean, default=False)  # 是否含 HowTo 格式
+    has_citations = Column(Boolean, default=False)  # 是否引用数据源
+    word_count = Column(Integer, default=0)
+    readability_score = Column(Integer, default=0)  # 可读性评分 0-100
+    heading_structure_score = Column(Integer, default=0)  # 标题结构评分 0-100
+    checked_at = Column(DateTime(timezone=True), server_default=func.now())
+
+    def to_dict(self):
+        return {
+            "id": self.id,
+            "article_id": self.article_id,
+            "topic_id": self.topic_id,
+            "platform": self.platform,
+            "total_score": self.total_score,
+            "has_schema": self.has_schema,
+            "schema_types": self.schema_types,
+            "has_faq_format": self.has_faq_format,
+            "has_howto_format": self.has_howto_format,
+            "has_citations": self.has_citations,
+            "word_count": self.word_count,
+            "readability_score": self.readability_score,
+            "heading_structure_score": self.heading_structure_score,
             "checked_at": self.checked_at.isoformat() if self.checked_at else None,
         }