Please have standalone documentation of modes such as autoAction

[qwertie]

This page says that when using makeAutoObservable, "All functions on prototype become autoAction." But what does this mean?

If one Googles "MobX autoAction" the top result is this page which does not appear to mention autoAction — well, actually it does, but a normal Ctrl+F won't find it! This is because the only mention of autoAction is hidden inside a tiny "note" which must be clicked to expand.

There should be a page specifically to document the modes available to makeAutoObservable (and to other functions that support autoAction), and the documentation of makeAutoObservable (et al) should link to it.

[qwertie]

It turns out that the documentation I was looking for already exists elsewhere on the page describing makeAutoObservable.

I would still suggest using links, and changing the description of autoAction from "methods that can act as action or derivation, based on their calling context" to "methods that can act as action or computed, based on their calling context" and then explain how MobX detects the "calling context".

[urugator]

action wraps the function with untracked to prevent accidental subscriptions. However in case of auto behavior, we don't know whether a function is intended for mutating state or as a helper method for reading state ("derivation") inside the tracking context (observer, autorun, reaction). So we introduced autoAction that allows both: it applies batching like action, it works with enforceAction etc, but it doesn't apply untracked, which allows for subscriptions. It never acts as computed.

Ragarding doc improvements, PR welcome :)