Add GT button to OCR overlay, prominent category picker, track pipeline

- Ground Truth button on last step of Pipeline/Kombi modes in ocr-overlay
- Prominent category picker in active session info bar (pulses when unset)
- GT badge shown when session has ground truth reference
- Backend: auto-detect pipeline from ocr_engine, store in GT snapshot
- Pipeline info shown in GT session list and regression reports
- Also pass pipeline param from ocr-pipeline StepGroundTruth

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

klausur-service/backend/ocr_pipeline_regression.py

@@ -12,7 +12,7 @@ import logging
 from datetime import datetime, timezone
 from typing import Any, Dict, List, Optional
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Query
 
 from grid_editor_api import _build_grid_core
 from ocr_pipeline_session_store import (
@@ -49,7 +49,10 @@ def _extract_cells_for_comparison(grid_result: dict) -> List[Dict[str, Any]]:
     return cells
 
 
-def _build_reference_snapshot(grid_result: dict) -> dict:
+def _build_reference_snapshot(
+    grid_result: dict,
+    pipeline: Optional[str] = None,
+) -> dict:
     """Build a ground-truth reference snapshot from a grid_editor_result."""
     cells = _extract_cells_for_comparison(grid_result)
 
@@ -57,9 +60,10 @@ def _build_reference_snapshot(grid_result: dict) -> dict:
     total_columns = sum(len(z.get("columns", [])) for z in grid_result.get("zones", []))
     total_rows = sum(len(z.get("rows", [])) for z in grid_result.get("zones", []))
 
-    return {
+    snapshot = {
         "saved_at": datetime.now(timezone.utc).isoformat(),
         "version": 1,
+        "pipeline": pipeline,
         "summary": {
             "total_zones": total_zones,
             "total_columns": total_columns,
@@ -68,6 +72,7 @@ def _build_reference_snapshot(grid_result: dict) -> dict:
         },
         "cells": cells,
     }
+    return snapshot
 
 
 def compare_grids(reference: dict, current: dict) -> dict:
@@ -160,7 +165,10 @@ def compare_grids(reference: dict, current: dict) -> dict:
 # ---------------------------------------------------------------------------
 
 @router.post("/sessions/{session_id}/mark-ground-truth")
-async def mark_ground_truth(session_id: str):
+async def mark_ground_truth(
+    session_id: str,
+    pipeline: Optional[str] = Query(None, description="Pipeline used: kombi, pipeline, paddle-direct"),
+):
     """Save the current build-grid result as ground-truth reference."""
     session = await get_session_db(session_id)
     if not session:
@@ -173,7 +181,18 @@ async def mark_ground_truth(session_id: str):
             detail="No grid_editor_result found. Run build-grid first.",
         )
 
-    reference = _build_reference_snapshot(grid_result)
+    # Auto-detect pipeline from word_result if not provided
+    if not pipeline:
+        wr = session.get("word_result") or {}
+        engine = wr.get("ocr_engine", "")
+        if engine in ("kombi", "rapid_kombi"):
+            pipeline = "kombi"
+        elif engine == "paddle_direct":
+            pipeline = "paddle-direct"
+        else:
+            pipeline = "pipeline"
+
+    reference = _build_reference_snapshot(grid_result, pipeline=pipeline)
 
     # Merge into existing ground_truth JSONB
     gt = session.get("ground_truth") or {}
@@ -224,6 +243,8 @@ async def list_ground_truth_sessions():
             "session_id": s["id"],
             "name": s.get("name", ""),
             "filename": s.get("filename", ""),
+            "document_category": s.get("document_category"),
+            "pipeline": ref.get("pipeline"),
             "saved_at": ref.get("saved_at"),
             "summary": ref.get("summary", {}),
         })