feat(data): integrate 9 UK government data sources via Kestra

Adds a full data integration pipeline for enriching school profiles with
supplementary data from Ofsted, GIAS, EES, IDACI, and FBIT.

Backend:
- Bump SCHEMA_VERSION to 3; add 8 new DB tables (ofsted_inspections,
  ofsted_parent_view, school_census, admissions, sen_detail, phonics,
  school_deprivation, school_finance) plus GIAS columns on schools
- Expose all supplementary data via GET /api/schools/{urn}
- Enrich school list responses with ofsted_grade + ofsted_date

Integrator (new service):
- FastAPI HTTP microservice; Kestra calls POST /run/{source}
- 9 source modules: ofsted, gias, parent_view, census, admissions,
  sen_detail, phonics, idaci, finance
- 9 Kestra flow YAMLs with scheduled triggers and 3× retry

Frontend:
- SchoolRow: colour-coded Ofsted badge (Outstanding/Good/RI/Inadequate)
- SchoolDetailView: 7 new sections — Ofsted sub-judgements, Parent View
  survey bars, Admissions, Pupils & Inclusion / SEN, Phonics, Deprivation
  Context, Finances
- types.ts: 8 new interfaces + extended School/SchoolDetailsResponse

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

backend/data_loader.py

@@ -16,7 +16,11 @@ from sqlalchemy.orm import joinedload, Session
 
 from .config import settings
 from .database import SessionLocal, get_db_session
-from .models import School, SchoolResult
+from .models import (
+    School, SchoolResult,
+    OfstedInspection, OfstedParentView, SchoolCensus,
+    SchoolAdmissions, SenDetail, Phonics, SchoolDeprivation, SchoolFinance,
+)
 from .schemas import SCHOOL_TYPE_MAP
 
 # Cache for user search postcode geocoding (not for school data)
@@ -381,6 +385,12 @@ def school_to_dict(school: School, include_results: bool = False) -> dict:
         "postcode": school.postcode,
         "latitude": school.latitude,
         "longitude": school.longitude,
+        # GIAS fields
+        "website": school.website,
+        "headteacher_name": school.headteacher_name,
+        "capacity": school.capacity,
+        "trust_name": school.trust_name,
+        "gender": school.gender,
     }
 
     if include_results and school.results:
@@ -455,8 +465,25 @@ def load_school_data_as_dataframe(db: Session = None) -> pd.DataFrame:
         # Query all schools with their results
         schools = db.query(School).options(joinedload(School.results)).all()
 
+        # Load Ofsted data into a lookup dict (urn → grade, date)
+        ofsted_lookup: Dict[int, dict] = {}
+        try:
+            ofsted_rows = db.query(
+                OfstedInspection.urn,
+                OfstedInspection.overall_effectiveness,
+                OfstedInspection.inspection_date,
+            ).all()
+            for o in ofsted_rows:
+                ofsted_lookup[o.urn] = {
+                    "ofsted_grade": o.overall_effectiveness,
+                    "ofsted_date": o.inspection_date.isoformat() if o.inspection_date else None,
+                }
+        except Exception:
+            pass  # Table may not exist yet on first run
+
         rows = []
         for school in schools:
+            ofsted = ofsted_lookup.get(school.urn, {})
             for result in school.results:
                 row = {
                     "urn": school.urn,
@@ -468,6 +495,15 @@ def load_school_data_as_dataframe(db: Session = None) -> pd.DataFrame:
                     "postcode": school.postcode,
                     "latitude": school.latitude,
                     "longitude": school.longitude,
+                    # GIAS fields
+                    "website": school.website,
+                    "headteacher_name": school.headteacher_name,
+                    "capacity": school.capacity,
+                    "trust_name": school.trust_name,
+                    "gender": school.gender,
+                    # Ofsted (for list view)
+                    "ofsted_grade": ofsted.get("ofsted_grade"),
+                    "ofsted_date": ofsted.get("ofsted_date"),
                     **result_to_dict(result)
                 }
                 rows.append(row)
@@ -511,3 +547,126 @@ def clear_cache():
     """Clear all caches."""
     global _df_cache
     _df_cache = None
+
+
+def get_supplementary_data(db: Session, urn: int) -> dict:
+    """
+    Fetch all supplementary data for a single school URN.
+    Returns a dict with keys: ofsted, parent_view, census, admissions, sen_detail,
+    phonics, deprivation, finance. Values are dicts or None.
+    """
+    result = {}
+
+    def safe_query(model, pk_field, latest_year_field=None):
+        try:
+            if latest_year_field:
+                row = (
+                    db.query(model)
+                    .filter(getattr(model, pk_field) == urn)
+                    .order_by(getattr(model, latest_year_field).desc())
+                    .first()
+                )
+            else:
+                row = db.query(model).filter(getattr(model, pk_field) == urn).first()
+            return row
+        except Exception:
+            return None
+
+    # Ofsted inspection
+    o = safe_query(OfstedInspection, "urn")
+    result["ofsted"] = {
+        "overall_effectiveness": o.overall_effectiveness,
+        "quality_of_education": o.quality_of_education,
+        "behaviour_attitudes": o.behaviour_attitudes,
+        "personal_development": o.personal_development,
+        "leadership_management": o.leadership_management,
+        "early_years_provision": o.early_years_provision,
+        "previous_overall": o.previous_overall,
+        "inspection_date": o.inspection_date.isoformat() if o.inspection_date else None,
+        "inspection_type": o.inspection_type,
+    } if o else None
+
+    # Parent View
+    pv = safe_query(OfstedParentView, "urn")
+    result["parent_view"] = {
+        "survey_date": pv.survey_date.isoformat() if pv.survey_date else None,
+        "total_responses": pv.total_responses,
+        "q_happy_pct": pv.q_happy_pct,
+        "q_safe_pct": pv.q_safe_pct,
+        "q_behaviour_pct": pv.q_behaviour_pct,
+        "q_bullying_pct": pv.q_bullying_pct,
+        "q_communication_pct": pv.q_communication_pct,
+        "q_progress_pct": pv.q_progress_pct,
+        "q_teaching_pct": pv.q_teaching_pct,
+        "q_information_pct": pv.q_information_pct,
+        "q_curriculum_pct": pv.q_curriculum_pct,
+        "q_future_pct": pv.q_future_pct,
+        "q_leadership_pct": pv.q_leadership_pct,
+        "q_wellbeing_pct": pv.q_wellbeing_pct,
+        "q_recommend_pct": pv.q_recommend_pct,
+        "q_sen_pct": pv.q_sen_pct,
+    } if pv else None
+
+    # School Census (latest year)
+    c = safe_query(SchoolCensus, "urn", "year")
+    result["census"] = {
+        "year": c.year,
+        "class_size_avg": c.class_size_avg,
+        "ethnicity_white_pct": c.ethnicity_white_pct,
+        "ethnicity_asian_pct": c.ethnicity_asian_pct,
+        "ethnicity_black_pct": c.ethnicity_black_pct,
+        "ethnicity_mixed_pct": c.ethnicity_mixed_pct,
+        "ethnicity_other_pct": c.ethnicity_other_pct,
+    } if c else None
+
+    # Admissions (latest year)
+    a = safe_query(SchoolAdmissions, "urn", "year")
+    result["admissions"] = {
+        "year": a.year,
+        "published_admission_number": a.published_admission_number,
+        "total_applications": a.total_applications,
+        "first_preference_offers_pct": a.first_preference_offers_pct,
+        "oversubscribed": a.oversubscribed,
+    } if a else None
+
+    # SEN Detail (latest year)
+    s = safe_query(SenDetail, "urn", "year")
+    result["sen_detail"] = {
+        "year": s.year,
+        "primary_need_speech_pct": s.primary_need_speech_pct,
+        "primary_need_autism_pct": s.primary_need_autism_pct,
+        "primary_need_mld_pct": s.primary_need_mld_pct,
+        "primary_need_spld_pct": s.primary_need_spld_pct,
+        "primary_need_semh_pct": s.primary_need_semh_pct,
+        "primary_need_physical_pct": s.primary_need_physical_pct,
+        "primary_need_other_pct": s.primary_need_other_pct,
+    } if s else None
+
+    # Phonics (latest year)
+    ph = safe_query(Phonics, "urn", "year")
+    result["phonics"] = {
+        "year": ph.year,
+        "year1_phonics_pct": ph.year1_phonics_pct,
+        "year2_phonics_pct": ph.year2_phonics_pct,
+    } if ph else None
+
+    # Deprivation
+    d = safe_query(SchoolDeprivation, "urn")
+    result["deprivation"] = {
+        "lsoa_code": d.lsoa_code,
+        "idaci_score": d.idaci_score,
+        "idaci_decile": d.idaci_decile,
+    } if d else None
+
+    # Finance (latest year)
+    f = safe_query(SchoolFinance, "urn", "year")
+    result["finance"] = {
+        "year": f.year,
+        "per_pupil_spend": f.per_pupil_spend,
+        "staff_cost_pct": f.staff_cost_pct,
+        "teacher_cost_pct": f.teacher_cost_pct,
+        "support_staff_cost_pct": f.support_staff_cost_pct,
+        "premises_cost_pct": f.premises_cost_pct,
+    } if f else None
+
+    return result