[FEATURE] Lint API specs

[dblock]

Is your feature request related to a problem?

In #203 we renamed a path parameter from NodeId to node_id because it caused a downstream generator in the opensearch-py client to produce a camelCase parameter which was caught in a linter there and errored.

What solution would you like?

A linter for consistency in the API spec for parameters, variables, etc.

Do you have any additional context?

opensearch-project/opensearch-py#700 (comment)

[nhtruong]

I think the reason that we should change params like NodeId to node_id is to stay consistent with the rest of the spec where snake case is used through out. While Python convention goes for node_id, other languages like JS actually prefers nodeId. The current JS client actually accepts both node_id (because this is what appears in the doc) and nodeId (because this is a JS' convention).

On a related topic, moving forward, we should get rid of nodeId in clients like JS. It adds overheads to the client code for a very small benefit. This will be breaking change though. So, we should combine this with the JS client rewrite in TS.

[nhtruong]

I will add this to the validator im working on as one of the follow up PRs.

[dblock]

Coming from #218, linter should check description format (capitalized, full sentences with periods), and no line wrapping.