add schema tests to the shell pipeline

[fauna-chase]

I validated that the task works as expected with fly execute. Once I verify it is working as expected in the pipeline I'll add it as a passed constraint to the publish job. I'll also move the integ tests to the test job, that way we will always auto run the tests and publish can be executed with the latest version that passed the tests (or any selected version that passed the tests)

[macmv]

looks good, except for the || comment.

might be nice to put this in the github PR ci too.

[macmv]

Suggested change

	if (collNames.length != expectedCollNames.length && !collNames.every((elem, idx) => elem === expectedCollNames[idx])) {
	throw new Error(`Schema collections do not match actual: ${collNames} expected: ${expectedCollNames}`)
	}
	const funcNames = await execPaginated("Function.all().map(.name)")
	if (funcNames.length != expectedFuncNames.length && !funcNames.every((elem, idx) => elem === expectedFuncNames[idx])) {
	throw new Error(`Schema functions do not match actual: ${collNames} expected: ${expectedCollNames}`)
	}
	if (collNames.length != expectedCollNames.length || !collNames.every((elem, idx) => elem === expectedCollNames[idx])) {
	throw new Error(`Schema collections do not match actual: ${collNames} expected: ${expectedCollNames}`)
	}
	const funcNames = await execPaginated("Function.all().map(.name)")
	if (funcNames.length != expectedFuncNames.length || !funcNames.every((elem, idx) => elem === expectedFuncNames[idx])) {
	throw new Error(`Schema functions do not match actual: ${collNames} expected: ${expectedCollNames}`)
	}

[fauna-chase]

looks good, except for the || comment.

might be nice to put this in the github PR ci too.

thats a good callout, I think I'll need to do a bit of tweaking there to use the docker image and create a fresh db and such so we don't run into any issues with them running concurrently. Definitely seems useful to get them on the prs in that manner.