fix: make migration runner more robust against existing schema elements

backend/run_migrations.py

@@ -44,7 +44,15 @@ def run_migrations():
             # Simple splitter by semicolon (works for basic schemas)
             statements = [s.strip() for s in sql.split(";") if s.strip()]
             for stmt in statements:
-                db.execute_commit(stmt)
+                try:
+                    db.execute_commit(stmt)
+                except Exception as e:
+                    # Ignore "Duplicate column name" or "Table already exists" errors
+                    err_msg = str(e).lower()
+                    if "duplicate column" in err_msg or "already exists" in err_msg or "duplicate key" in err_msg:
+                        print(f"  Note: Statement skipped (already applied or exists): {stmt[:50]}...")
+                    else:
+                        raise e
             
             # Mark as applied
             db.execute_commit("INSERT INTO schema_migrations (migration_name) VALUES (%s)", (filename,))