What prevents the reworked draconians from attacking other civs?
There is no definitive answer, really. When a civ evaluates whether they are willing to attack or not, there are a lot of things that play a role.
First is ethics. The ones you have here make draconians hated by elves and humans, thus, they probably get attacked by them a lot. They don't hate humans and elves back
as much, because it's mostly the draconians who do what is taboo for humans/elves - not the opposite way around. Judging by some of my tests, I also suspect that a "hate overflow" is possible (too much hate turns into very mild hate after a certain number is reached), but that's a wild guess. Anyhow,
the table here is helpful for calculating animosity. Note that
only certain ethics are taken into consideration for a war.
Second is population. Got a smaller population than the target civ - is less likely to attack the target civ.
Third is the ruler's values. An unlikely culprit here, though, with such a low value for peace.
Fourth is triggers for war. There are some random, out-of-nowhere events in worldgen for this, but ITEM_THIEF and BABYSNATCHER are more steady sources of such triggers. In my tests, BABYSNATCHER almost always made a civ more warlike, so it seems to be the stronger one.
Fifth is the site type shenanigans. For unknown reasons, civs with CAVE_DETAILED tend to be more passive. May also be the case here, with such a weird site type. Try out CITY and see if it makes a difference.
Sixth is borders and competition. Isolated civs and civs that have room to expand usually were less warlike in my tests. They just keep peacefully expanding in their permitted biomes, uncontested.
Sixth is pure luck. Sometimes a civ just ain't willing to do anything. Small ethic alterations here and there usually helped me to push the civ out of such moot point.