> how does it provide a sanity check between what the back end delivers and the front end expects
The team I work on have been using Mirage with GraphQL for 6 months or so and GraphQL’s tooling almost entirely addresses this question. There will still be false positives in API behaviour (you can never have a completely accurate simulation) but you can get an incredible degree of confidence at a startlingly low cost, with a delight developer experience to boot. Similar strategies could no doubt be employed for JSONSchema, Swagger, or Pact tooling.
The team I work on have been using Mirage with GraphQL for 6 months or so and GraphQL’s tooling almost entirely addresses this question. There will still be false positives in API behaviour (you can never have a completely accurate simulation) but you can get an incredible degree of confidence at a startlingly low cost, with a delight developer experience to boot. Similar strategies could no doubt be employed for JSONSchema, Swagger, or Pact tooling.