Version 1.2.3
– Added some differences in building priorities based on country situation. For example, resource extraction is now a bit more important for subjects and poor countries compared to developed nations.
– Improved support of compatibility patches. Shared building limits are now checked dynamically, allowing the system to work with modded buildings. The script will actually choose between oil rigs and synthetic oil instead of just building both.
– AI now always balances the budget as if construction sectors were fully loaded, leading to much better results for countries that are in debt.
– Government and military building expenses are now calculated using the average instead of the highest value from the previous month.
– Adjusted how budget surplus is calculated when reserves or credit are full.
– AI is now better at recognizing when a debt is dangerous enough for it to start downsizing government and military buildings.
– Improved how AI determines which production buildings are excessive.
– Excessive production buildings are now integrated into calculations of available workforce in a state.
– Further increased priority of some government and military buildings.
– Small improvements of building script and a couple of bug fixes.
– Reduced performance tax of compatibility patch system.
– Removed unused modifiers.
– Removed changes to decrees.
– Removed changes to government administration building.

Version 1.2.2
– Compatibility with patch 1.1 changes.
– Massive code refactoring to separate building data from the script itself. Hopefully without breaking anything in the process... This will allow for a proper support of long awaited compatibility patches, so you will be able to enjoy both custom buildings and good AI at the same time.
– Allowed AI to raise government and military wage levels to high and very high.
– AI countries won't be forced to research Pumpjacks before they have Rubber Mastication anymore.

Version 1.2.1
– Construction game rules now also support positive values up to 150%. Warning: everything over 100% give construction efficiency buffs to AI countries, so it's no longer fair play. These options are meant for players who just want AI to develop even faster and don't mind giving them advantage to achieve this.
– Refactored how budget surplus gets calculated to improve accuracy and fix incorrect results for countries in debt.
– Increased priority of all government and military buildings, resulting in targets being followed more consistently, e.g. countries like Russia actually hitting innovation cap.
– Adjusted building targets.
– Improvements to the algorithm that controls the level of taxes and government wages.
– Tweaked budget health check.
– AI countries are now less aggressive in downsizing construction sectors.
– Added a fix for strike negotiations where you could receive demands to downgrade labor rights.

Version 1.2.0
– Added game rules that will allow you to set the amount of construction points wanted by AI countries in the range from 100% to 50% with a separate value for subjects and junior market partners of player countries, replacing ARoAI Difficulty game rule that had only a few options to choose from. It's not only about the ability to make the game easier now, but also about adjusting the pace of the playthrough and your role in it, as fewer construction points means AI countries won't develop as fast as usual. Should be helpful if you think AI is too perfect with the mod (like a certain One Proud Bavarian does) and want some middle ground.
– Military targets got some additional factors affecting them. Desired amount of battalions and flotillas now also scale with how weak the country is compared to potential threats it may face, which mostly affects small nations. At the same time, the amount of money that is allowed to be spent on the army and navy now depends on the difference between the desired and the current number of battalions and flotillas, preventing countries with temporary budget issues from becoming defenseless. What is more, if AI country is already much stronger than basically everyone else, it will reduce amount of required barracks and naval bases to save resources. It is also important to note that all these calculations take into account the quality of the country's army and navy, as well as how many battalions it can conscript.
– Government and military building expenses are now processed separately for each type of building instead of being grouped up. This will allow AI countries to control their budgets much more precisely.
– Downsizing algorithm was modified to support changes to calculations of expenses with some additional general improvements as well.
– Yet another rework of how building targets scale with GDP per capita. Adjusted targets themselves as well and swapped some of them from being based solely on incorporated population to partially consider unincorporated population too.
– AI countries will now slightly exceed building targets if they can afford it.
– Amount of workforce that is available in a state is now calculated in a much more accurate way.
– Fixed vanilla issue of AI not using advanced production methods in universities.
– Increased priority of using advanced production methods in government administrations.
– Allowed AI countries to use automation production methods that require oil.
– AI will now try to build at least a few oil rigs to see if there's a demand for more.
– Increased priority of opium plantations.
– Fixed some problems regarding construction of railways and ports.
– Improvements of decision making on building government administrations in countries with tax capacity issues.
– Added Skyscraper and Mosque of Djenne to the script.
– Building cooldowns and other similar checks now use hidden timed variables instead of modifiers that are visible in UI.
– Slightly reduced said building cooldowns.
– Fixed wartime spending mode lasting much longer than intended.
– Some building conditions are now considered critical and will bypass limits like being in debt, for example country having negative bureaucracy or lacking port connections.
– Prohibited AI countries from downsizing government and military buildings during heavy bureaucracy shortages to force them into fixing it first.
– Adjusted a couple of defines to increase stability of consumption taxes usage by AI.
– Fixed calculation of budget surplus not working if gold reserves are full.
– Countries with no construction sectors will now ignore debt when determining if they want to queue new buildings as construction is free for them.
– Autobuild won't lower amount of buildings it wants to queue during debt anymore.
– Fixed expected costs of new construction sectors being too high under certain circumstances.

Version 1.1.1
– Fixed non-English localization not working.
– Adjusted basically everything about universities.
– Tweaked how building targets scale with GDP per capita.
– AI won't build 20 Suez Canals anymore.

Version 1.1.0
– Autobuild decision now opens a menu where you can not only enable or disable the system, but also choose which government and military buildings you allow it to construct.
– Added a new game rule to adjust difficulty level. "Easy" option will allow you to slow down economy growth of AI countries, making them weaker. There's also "Easy Except Subjects" option that works the same as "Easy", but it won't apply to subjects and junior market partners of player countries.
– AI now takes into account potential budget expenses of new government and military buildings before constructing them, which is very important in cases like small countries getting their first few construction sectors.
– Changed how income values are evaluated when setting spending limits, which should improve budget stability in AI countries. Spending limits were adjusted accordingly.
– Downsizing algorithm should now work better in small countries and in critical situations in general.
– AI now queues fewer government and military buildings at a time to reduce overbuilding of them.
– Target construction points are now calculated using total population of the country, rather than excluding half of unincorporated population.
– Improved how AI determines if there is enough available workforce in a state.
– AI won't try to fix tax capacity in unincorporated states anymore.
– Buildings that are not present in the country yet are now constructed in a random state with high available workforce rather than strictly in a state where it is highest.
– Budget income calculation now takes bankruptcies into account.
– AI now calculates excess values of government and military buildings more accurately.
– Increased importance of consumption taxes for AI.
– Tweaked thresholds of budget health check.
– Lowered target convoys for isolationist countries.
– Fixed AI ignoring different types of luxury drinks being substitutes to each other.
– Removed vanilla priorities of some buildings to stop AI from moving them to the top of the queue.
– Hardcoded automation production methods are now a bit less hardcoded.
– Improvements in the code of the research simulation system.
– Fixed custom 0 war support fix causing revolutionary and secessionist countries to become invincible to plays under certain circumstances.

Version 1.0.7a
– Ruled out a potential variable overflow.

Version 1.0.7
– Fixed typo found by dgsaf from Victoria 3 subreddit that caused AI to almost always think it has a powerful protector in protectorate, tributary and violate sovereignty checks, noticeably lowering acceptance scope for no good reason.
– Removed changes to infrastructure requirements of trade sectors, as this issue is now coreved by an official hotfix.
– Vanilla building algorithm is now properly suppressed. Previously it could still a bit of government buildings to the queue, creating budget instability.
– Fixed government administration building not being constructed by AI under certain circumstances.
– Tax level switching is now a bit more flexible.
– Fixed minor bugs in building algorithm.
– Lowered frequency of AI spending update, as it becomes less important when construction is controlled by custom script.

Version 1.0.6
– Ensured compatibility with patch 1.0.4 changes like diplomacy and warfare AI improvements, Petite Bourgeoisie caring about government wages, training rate debuff of military wages and typos in defines being fixed.
– [GAME BALANCE] Removed infrastructure requirements of trade centers added in patch 1.0.4, as this change is straight up game breaking and is confirmed to be unintentional by devs in its current state.
– Pumpjacks, a technology behind discovering and extraction of oil, is how prioritized by AI when it becomes available.
– Fixed downsizing of production buildings being effectively disabled in most cases due to too harsh conditions. Also adjusted the algorithm itself.
– High GDP per capita scaling of various building targets is changed to support further growth of the richest countries.
– Changed navy size targets, raising average value and lowering difference between mostly coastal and mostly continental countries.
– Improvements to how building targets interact with current tax level.
– Adjusted downsizing conditions for non-military government buildings.
– Construction of buildings by AI and Autobuild now starts in the first month of the game instead of second one.
– Preferred infrastructure when choosing states for crically important buildings is now higher.
– Minor performance improvements in some systems.

Version 1.0.5
– Arable resources data is now stored in a much more efficient way of using memory, potentially improving stability of the game. This change is save compatible, data in your existing games will be recreated.
– Market capital connection checks from the previous update turned out to be quite heavy in terms of performance. Optimized the most problematic parts and fixed a few bugs here, should be relatively fine now.
– Further improvements to downsizing algorithm flexibility and priorities.
– Military targets are less now dependant on GDP in general, but ceiling is also noticeably higher, allowing rich countries to have an edge.
– Prussia got a bit higher than usual priorities on how much army to keep in the early game, as their needs are justified by historical position more than by actual economic capabilities.
– Target convoys now scale further with country's GDP per capita instead of peaking at medium values.
– Tweaked conditions of building railways.