> 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.
I hope the post is fair. Let me know if it isn’t. Our goal is to explain the situation to our customers, and help make other teams aware that their apps may be suffering similar issues.
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.