I've only ever seen this message when trying to grant a title to someone who is unlanded (eg. monk, celibate). In this situation, you can work around the problem by first granting them a temple holding somewhere (preferably a temple that already exists in the county), and then you grant them the county title.
check the file in \common\opinion_modifiers\ for that, I altered that opinion modifier as well. For the -60, you want
opinion_revoked_my_title. You may also be interested in opinion_retracted_my_vassal and opinion_retracted_my_de_jure_vassal. Offhand I don't know the one for the -15 opinion modifier from other vassals, but I'm sure it's in there somewhere.can_be_spiritual_trigger in \common\scripted_triggers\00_scripted_triggers.txt.Thanks very much for the answers and regarding the last question: Do you know if it is possible to add inability to marry (like bishops for example) as one of the conditions to be spiritual leader/chaplain ? Something like "can_marry = no" or "can_get_married = no" ? Does something like that ever exist in triggers? Because I know there is "is_married = yes/no" but it refers only to specific moment, not inability to marry at all.
can_marry = no is what you want, according to the wiki here https://ck2.paradoxwikis.com/Conditions
Exactly what I want. I'm trying to add possibility for custom playable theocracy government to be court chaplain (because game apparently detects it as entirely different government from feudal group) so I thought about somehow combining the following into one trigger: can_marry = no, is_christian = yes (to include only clerical bishop celibacy) and probably holy_order = no because as far as I know they can't marry either just like bishops. Though I wanted to add these making it possible to become chaplain aside from other options and not absolutely required (if it's possible at all) like i.e. independent = no.
preferred_limit, on careful reading, seems to suggest that the game will stop looking for matches once it meets a preferred limit that qualifies. However, the example used there is for a clearly nested set of preferred limits, that is, the first of three preferred limits in the example is is_powerful_vassal = yes, followed by being a tier of KING or DUKE, followed by a tier of COUNT.random_vassal = {
limit = { trait = ambitious }
preferred_limit = { is_powerful_vassal = yes }
preferred_limit = { trait = diligent }
}The former. You need to use more thorough preferred_limits to cover several preferences, e.g.
random_vassal = {
limit = {
trait = ambitious
}
preferred_limit = {
is_powerful_vassal = yes
trait = diligent
}
preferred_limit = {
is_powerful_vassal = yes
}
preferred_limit = {
trait = diligent
}
# Commands
}From bitter experience, I'll add that this approach becomes very tedious very quickly as the number of conditions increases.
preferred_limit clauses for n conditions. 2 conditions => 3 clauses; 3 conditions => 7 clauses; 4 conditions => 15 clauses; etc. But that's just the raw number of clauses - it says nothing about the headaches from ordering. Eg: Is diligent+king better or worse than powerful_vassal+duke?)This certainly explains why my code wasn't fully functioning as intended. I'm going to have to touch up so much of my code now.
on_host_change fire for landed characters, for example if they're taken prisoner?Have you considered testing it? (Eg:
on_host_change, fire a hidden event on the character, which fires a visible event on their current host. Then imprison a landed character with the console - do you receive the visible event?) And then, if the wiki's unclear/misleading, you can update it 
Have you considered testing it? (Eg:on_host_change, fire a hidden event on the character, which fires a visible event on their current host. Then imprison a landed character with the console - do you receive the visible event?) And then, if the wiki's unclear/misleading, you can update it
That sounds like your imprisonment action did not change the character's host (for the purpose of the on_action, at least), but something else did a few months later.
I actually let the game run a little bit, to see if anything would fire the test event. During the process, I noticed that a vassal who my character was guardian of, to educate a matching culture for in order to get rid of the opinion malus, was given another guardian in his own court, and concluded that the on_action didn't work. Then, a few minutes later, but a few months longer in the game, I got the test event firing to notify me of that same vassal changing host, which is what confused me.