Still not 100% clear on what you're getting at.
Maybe we can take a simple scenario like the one below, and analyze things from there.
*
Two dominos, A and B and an agent, X.
X pushes Domino A, causing Domino A to fall against Domino B, causing Domino B to fall.
Domino B falling:
Proximal cause = Domino A falling against it
Distal (ultimate) cause = X pushing Domino A.
*
I think “distal” is a better term than “ultimate” because ultimate causes are never really ultimate, and are always also proximal to some effect in a chain.
Anyhow, causal network and contextual orientation ought to be part of the meta-context here in terms of placing the question in a comprehensible form.
Consider the causal network as the network of enabling conditions and triggering causes that could possibly be considered relevant in the outcome, such that proximal cause can be defined with some flexibility. This causal network causes a kind of fuzziness around the identification of the proximal cause.
Note this network as a whole can be considered the necessary and sufficient conditions of the effect under analysis. Also, the proximal cause is usually a triggering cause (and necessary e.g. my push) rather than an enabling condition (maybe only contingent, depending on the circumstances, e.g. the weight of the domino).
Re contextual orientation, are we approaching this from the context of a human observer with human desires, needs, goals, and recognizably human actions? Are we focusing on the mechanics of the situation? The physics? E.g. perhaps down to the micro context of electrons? The neurobiology? Our analysis can be psychological, chemical, physical or some combination thereof depending.
All relevant considerations.