Much of indexing strategy depends on knowing the data like how many distinct values and what the distribution is like. Is JsonBField->>'status' always set? Are those three values mentioned in this query common or rare? Can you re-write this query to avoid using an OR in the where clause? Are you just wanting to add a GIN index for the jsonb paths? Or do you want indexed like below that are a bit stylized to this query?
CREATE INDEX idx_MyTable_status USING btree( JsonBField->>'status' );
CREATE INDEX idx_MyTable_descartada_date USING btree( To_Date(JsonBField->'descartada'->>'data','yyyy-mm-dd') );
CREATE INDEX idx_MyTable_contrato_date USING btree( To_Date(JsonBField->'contrato'->>'data','yyyy-mm-dd') );
