Spies are vastly overpowered when used to open gates for an assault. Even a newborn spy, from a brothel in a city with no guild, sent to a rebel settlement, can have 40% or more chance to open the gates. Worse, even if he fails, the game recalculates the chance on the following turn, so you're almost gauranteed a successful result within a few turns. Considering the goals of TLR, this seems even more out of place than in vanilla or other mods.
1) Would really like to see the % chance changed to, say, 1% per skill lvl of the spy minus 1% per lvl of the settlement who's gate he's trying to open. I doubt you'll be able to do this, but IMO this would give a much more historical success rate.
2) If you can't change the success rate, would it be possible to add a trait line for spies that reduces their success rate for opening gates? Something like, a -x.95 chance of success trait that 100% of spies get at birth (thereby reducing that 40% chance to 2%). Could then have a line of traits in sequence, progressively more difficult to acquire, that have less of a reduction in success chance?
3) If you can't do a slick trait line to get the results desired, is it at least possible to give all spies at birth a trait/ancillary that completely destroys any chance that they'll ever open a gate?
As currently implemented, players have only 2 choices: use the spies and expand at a ridiculous rate thereby pre-empting AI faction expansion opportunities OR implement a House Rule to never use spies to open gates, utilizing self-control to make the game more challenging. If you could mod them to something more reasonable, players could actually use them and not get a ridiculous result.