feat(group-a): Foundation for Fairthon 75-Technique system (#224, #225, #226, #227)

- #224: Remove input_source field, add fairthon_source for YAML preservation
  - Remove filter_techniques() and determine_available_inputs() from loader
  - Add get_all_techniques() and has_readme_content() to registry

- #225: Replace 75 Korean YAML technique definitions with English Fairthon templates
  - Map Fairthon categories to wine-themed folder names
  - Remove nameKo field from schema and all YAMLs

- #226: Unify EvaluationMode enum as single source of truth
  - Remove duplicate EvaluationMode from models/graph.py
  - Replace SIX_HATS with SIX_SOMMELIERS everywhere
  - Add FULL_TECHNIQUES mode
  - Update frontend TypeScript types

- #227: Define BMAD 17-item evaluation canon (A1-D4, 100 points)
  - Create backend/app/criteria/ module with bmad_items.py
  - A=25pts, B=25pts, C=30pts, D=20pts

- Fix pre-existing auth test failures (7 tests)
  - Update _check_ownership tests for 3-param signature
  - Fix Secure cookie assertion for localhost environment

Closes #224, Closes #225, Closes #226, Closes #227

Author: ComBba

backend/app/api/routes/graph.py

@@ -21,15 +21,15 @@
 }
 from app.core.exceptions import CorkedError, EmptyCellarError
 from app.database.repositories.evaluation import EvaluationRepository
+from app.graph.graph_factory import EvaluationMode
 from app.models.graph import (
     ReactFlowGraph,
     TraceEvent,
     ModeResponse,
-    EvaluationMode,
     Graph3DPayload,
 )
 from app.services.graph_builder import (
-    build_six_hats_topology,
+    build_six_sommeliers_topology,
     build_full_techniques_topology,
 )
 from app.services.graph_builder_3d import build_3d_graph
@@ -66,11 +66,11 @@ def _check_ownership(evaluation: dict[str, Any], user, evaluation_id: str) -> No
     # Allow access to public demo evaluations without auth
     if evaluation_id in PUBLIC_DEMO_EVALUATIONS:
         return
-    
+
     # Require auth for non-public evaluations
     if user is None:
         raise CorkedError("Authentication required to view this evaluation")
-    
+
     if evaluation.get("user_id") != user.id:
         raise CorkedError(
             "Access denied: evaluation belongs to another user", status_code=403
@@ -84,14 +84,16 @@ def _determine_mode(evaluation: dict[str, Any]) -> EvaluationMode:
         evaluation: The evaluation document.
 
     Returns:
-        EvaluationMode (six_hats or full_techniques).
+        EvaluationMode (six_sommeliers, grand_tasting, or full_techniques).
     """
-    # Check for explicit mode in evaluation data
-    mode = evaluation.get("mode")
+    # Check for explicit evaluation_mode first, then fall back to mode
+    mode = evaluation.get("evaluation_mode") or evaluation.get("mode")
     if mode == EvaluationMode.FULL_TECHNIQUES.value:
         return EvaluationMode.FULL_TECHNIQUES
-    # Default to six_hats mode
-    return EvaluationMode.SIX_HATS
+    if mode == EvaluationMode.GRAND_TASTING.value:
+        return EvaluationMode.GRAND_TASTING
+    # Default to six_sommeliers mode
+    return EvaluationMode.SIX_SOMMELIERS
 
 
 @router.get("/{evaluation_id}/graph", response_model=ReactFlowGraph)
@@ -133,7 +135,7 @@ async def get_graph(
     if mode == EvaluationMode.FULL_TECHNIQUES:
         graph = build_full_techniques_topology()
     else:
-        graph = build_six_hats_topology()
+        graph = build_six_sommeliers_topology()
 
     logger.info(f"[Graph] Returning {mode.value} graph for {evaluation_id}")
     return graph
@@ -160,7 +162,7 @@ async def get_graph_structure(
     if mode == EvaluationMode.FULL_TECHNIQUES:
         graph = build_full_techniques_topology()
     else:
-        graph = build_six_hats_topology()
+        graph = build_six_sommeliers_topology()
 
     return graph
 
@@ -186,7 +188,7 @@ async def get_graph_execution(
     if mode == EvaluationMode.FULL_TECHNIQUES:
         graph = build_full_techniques_topology()
     else:
-        graph = build_six_hats_topology()
+        graph = build_six_sommeliers_topology()
 
     methodology_trace = evaluation.get("methodology_trace", [])
     if methodology_trace:
@@ -268,7 +270,7 @@ async def get_mode(
 ) -> ModeResponse:
     """Get current evaluation mode.
 
-    Returns the evaluation mode (six_hats or full_techniques)
+    Returns the evaluation mode (six_sommeliers, grand_tasting, or full_techniques)
     for the specified evaluation.
 
     Public demo evaluations can be accessed without authentication.

backend/app/core/config.py

@@ -38,6 +38,7 @@ class Settings(BaseSettings):
 
     # LLM APIs
     GEMINI_API_KEY: str = ""
+    OPENAI_API_KEY: str = ""
 
     # Vertex AI Express (API key auth)
     VERTEX_API_KEY: str = ""

backend/app/criteria/__init__.py

@@ -0,0 +1 @@
+"""Evaluation criteria module for BMAD 17-item scoring system."""