[experiment] See if examples (from the docs) is all you need

[mikecann]

I posted this to Slack too but here is a thought that I would love to know the answer to.

Rather than writing individual guidelines explaining how to use Convex we instead just provide a number of examples as guidelines for the AI to use.

We could source these from the Convex docs that way it could be guaranteed to always be up to date as the docs should always reflect the latest APIs.

It might be interesting to run this as an experiment.

I'm leaving it here to see if anyone else wants to jump on it, if not ill take a look when I get a min :)