Every other passive affinity bonus seems to break game rules or limits in some way. They feel pretty awesome, allowing for a distinctive playstyle:

Scientists - Can research one era ahead; 

Expansionist - Can trespass without open borders; 

Militarist - Can declare formal wars without grievances/demands; 

Merchant - Can resell luxuries;

Aesthete - Maxed out cultural proximity.

But then you have the Builders and Agrarians which read:

Builder - Gains 10 Stability when completing a District's construction.

Agrarian - Gains 5 Stability when gaining Population.

This feels a bit dull since they don't quite break the rules like the remaining affinities and are too similar to one another. I have a couple of suggestions, but I apologise in advance since I'm new to the game and I'll likely be missing the full implications of my suggestions. In any case, these are:

Builder, Pride: Gains 10 Stability when completing a District. Can accrue up to 150% Stability. 

100% to 150% Euphoric: +3 Influence per Population (rather than +2 from Settled).

Lifting the Stability limit allows the Builder culture to take greater advantage of its passive ability. It's still very close to how it currently works, but it adds greater incentive to take full advantage of it and it adds an element of uniqueness to its passive which it currently lacks.

Breaks the Stability rule (Stability rule: can't go over 100%).

Agrarian, Enclosures: May exploit Food in tiles adjacent to District but outside its territory, including foreign territory.

This gives Agrarian cultures greater ease in the placement of Districts which exploit Food tiles. It also increases the value of border tiles for the purpose of district placement.

Breaks the Territory rule (Territory rule: can't exploit tiles outside District's territory).

---

What other ways do you think this could be addressed? I think it would be fine for a single cultural affinity to have a very basic and simple passive, so either one of these could just remain as it is (probably Agrarian). Having different levels of complexity is probably a good thing. But what bothers me is that both passives are essentially the same, and I think one should be changed and work more like the other remaining 5.