Add response vs model / spec validation

[black-snow]

Is your feature request related to a problem? Please describe.
Either I'm blind and cannot find this or I must be awkward - this feature seems to be quite essential to me (and pretty much half of what assertible.com does):
I'd like to be able to validate my requests from the DEBUG view against my model spec defined in the DESIGN view. And/or I'd like to be able to automatically set up tests for my spec in the TEST view. How come nobody has requested this yet?

Describe the solution you'd like
When I hit the DEBUG tab insomnia creates my paths/requests for me, great! ('cept for some minor issues). But how can I check if the responses I receive actually validate against the model/spec? This should be done automatically or at least be reachable with one click.
When I hit the TEST tab insomnia presents me with the requests I specified in the DEBUG tab, great! But why aren't there tests to check against my spec by default? They should be generated automagically or be reachable with one click.

Describe alternatives you've considered
Alternatively I could manually write a test expect(body).to.be.an(xyz) ... until I have manually recreated the whole model spec ... :(

Additional context
Atm I'm reverse engineering an API so I write the spec against the responses I see. After I'm confident with what I wrote I'd like to validate that my spec actually matches the response(s). And in the last step I'd like to generate client code / SDK to use in my application.

[nijikokun]

Absolutely.

When I hit the TEST tab insomnia presents me with the requests I specified in the DEBUG tab, great! But why aren't there tests to check against my spec by default? They should be generated automagically or be reachable with one click.

We had it to be slated as a part of the unit testing feature-set but was cut from scope due to timing. We haven't really had time to build it out yet, the idea was to take the spec and generate unit tests based on it for the user like you mention under the alternatives so you wouldn't have to. Would that be a viable solution?

[black-snow]

Hey, thanks for your quick reply. That sounds like a solution - as long as the error reporting is good ;) If you take a look at assertible.com for example, the reporting is pretty bad imho, like "something is not good somewhere over there". But a very good solution will probably be a huge pile of work, both technically and on the UX side.

As a user I'd expect the tool not just to tell me the json path to where it failed but to show me in the spec what was expected and what I actually got. (Ideally even with suggestions how to fix it - like "you specified string but got null, add null/nullable here?")
So a side-by-side view could be good, spec on the left, response on the right - somewhat like a two-way merge window. Or maybe three-way with the middle being a playground where you can modify your model and apply suggestions in order to make it match the response. (Sure, usually the model should be there first and the response should match the model spec, not the other way around.)

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[black-snow]

not stale