Bugfixes:

- Now AI buying heroes on market checks if the Isander's gambit competitive quest have the marketplace heroes locked.

- AI will not try and fail to assign a hero to a system that is blockaded.

- Fixed an instance where Lumeris could not buyout uncolonized systems under their influence radius.

- Added a Minimum fitness to all fleet request, to avoid military fleets slipping into specialized missions (explore, colonization, etc).

- Adjusted Disband Action to not call the core system if the resulting ship list is empty.

- Fixed an error with intercept fleet quest assignation.

- Fixed formulas for Nakalim Relic usage to avoid assigning relics when they have very few relics (and errors trying to do that...)

- Fixed Horatio trying to Gene splice over and over (generating lots of errors) in the same turn it splices successfully.

- Fixed outpost migration trying to change food envoys source when outpost if full or will grow next turn (and generating lots of errors).

Improvements:

- AI will assign their governor heroes to their most important system available instead of just in system acquisition order.

- New generator for Ship Bound's Essence to modify interest in using Holy Proliferation (overuse instead of expanding) and Alms for essence (no use at all, now should be rare if conditions are met).

- Lowered the interest a bit for researching techs with Ark modules.

- Minor adjustments in Leecher and Invader fleet size progression.

Heroes and Nakalim

On Hero marketplace purchase and selection from events or academy, AI will consider if the hero is of a missing political party, and specially a missing senate leading party.

Nakalim fleet commander heroes will also be able to use the relics as hero skill.

Nakalim will stop assigning relics to heroes if more than one third of the total relics are assigned to them.

Diplomacy

- AI will reset their diplomatic desire as expected in each turn analysis computation.

- AI will not try and fail to send a "Bureaucratic Imbroglio" if it is already running.

- AI will not try and fail to send a Science Agreement if it doesn't have the Alliance Ability requirement.

- AI will not try and fail to propose peace or alliance if any side has the eternal war faction trait (Cravers).

- AI will not spam alliance request every turn (plus lots of extra internal tries and fails), will be restricted to one negotiation request each two turns, as it was originally programmed.

Bugfixes:

- Fixes a possible instance the Small Invader design could lose his role.

- Extended the bugfix for Lodestone systems to also cover rejuvenated systems.

- AI will not try to build a behemoth when other behemoth is already being built in their systems.

- Fixed generator for queuing buildings for max population on planets, now has the correct parent assignation, also half interest if there are colonizable planets in the system.

- Fix check of approval resulting of planet colonization against empire approval instead of system approval.

Improvements:

- New generator for technology interest for all the empire upgrades and buildings for extra population slots in planets.

Bugfixes

- AI Nakalim is now able to build temples (was missing a resource allocator for empire relics!).

- AI will not try and fail to colonize Lodestone Systems (Revealed by Isander's Gambit competitive quest).

- AI will not try to assign a hero to the Academy Master Fleet.

- Nakalim AI will not try (and fail and leave the system queue empty) to make Cathedral to the Lost in other systems while one is being built.

- AI will not try to disband Behemoths into the hangar.

Improvements

- AI Nakalim Temple placement is improved, will build them where most FIDSI will be gained from the temple, and will spend one third of their total relics in temples.

- Added an new generator: AI Heroes will prioritize assigning senator skills to heroes aligned with their current leading political parties.

Bugfixes:

- Fixed the possibility for AI to not queue anymore in systems with infinite buildings.

- Fixes on calculations of wanted strategic resources, didn't account properly for strategic modules.

- AI will not try to propose diplomacy contracts to eliminated players.

- AI will not try and fail to propose cooperation agreement if they don't have the ability.

- AI will not ask for a trade/science agreement if they initially can't afford the costs.

Missing interests in Economy And Trade:

- Added proper interest to the hero/ship marketplace unlock, based on the Empire Dust ability (dust stock + production)

- Added minor interest to buyout reduction.

- Interest planet improvements that increase resource extraction now is based on the resources need and related deposits owned.

- Interest in everything related to trade routes is based in both number of systems owned and exploited luxuries that the AI needs.

- Added missing interest for trade route improvements.

Marketplace:

- AI will be more thoughtful about what resources to stock:

If needs several system upgrades, projects their production.
Detects bottlenecks in a recipe (based on current stock and production of that luxury) and lowers a bit the amount wanted for the other luxuries of that recipe.
Stock a bit of additional luxuries for future upgrades.

- Fixed the relative value of a resource in the market being directly attached to their base value.

- Now AI takes into account manpower upgrades and if uses strategics for system upgrades when saving strategics.

Technologies

- Interest to unlock and improve trade routes now raises a bit with the number of colonies.

- Increase luxuries and strategics has now a secondary minor source of interest for developing systems, aside the ensure enough taskforce one (meant for strategics only).

- Interest for curiosity detection also affects level 3 and 4 curiosity science improvements, AI players with restless personality has a bit more interest.

- AI will not consider unique buildings, wonders, planet improvements, infinite buildings etc as pending to build when decreasing the interest of researching a tech for resource increase.

- Adjusted the way the individual resources compound into technology interest, as the way is was calculated, were heavily penalized.

Other:

- AI will have much less interest in the victories (that affect some things like interest in ship improvements, or interest in dust) based on the number of technologies researched, 50% less interest with less than 20 techs (early game) and caps out at full interest with 40 technologies or more (mid to late game).

- Added default victory choices for Hissho (Supremacy) and Umbral Choir (Economic).

- Fixed the analysis job with the number of estimated wanted systems used for colonization and expansion, now they exclude the special nodes and the academy owned nodes for the computation. Also takes in to account the game speed.

- Adjusted and fixed the mess of the colonization interest based on wanted systems, boosted interest for Imperious personality, takes into account outposts at a reduced value.

- Now AI really chooses the best resources for system upgrade based on the value of the ingredient and will wait until has enough resources to build them. Before AI would choose the three first with positive production.

- AI will prioritize resource planet improvements (IE: Slag and Sludge Center) on planets where are resources that the AI wants, and blocks other planet improvements while still wants the resource.

Finally fixed properly the retreat function. I messed it up a bit previously.

Leecher fleets will not try (and fail) to leech systems of adversaries with closed borders, or the academy (is not possible, even at war).

Optimizations in fleet allocators, takes slightly into account distance to target in order to choose one fleet or other, not only fleet fitness for the task.

Improvements in resource needs. Some values disappear as it recalculates over the turn, so must be retained with internal variables. Lowered a tiny bit the research and dust base interest.

Diplomacy:

AI will not inflate their military power to simulate aggressiveness so much.

AI will lean towards a cordial cold war state if enough time passed since entered cold war and nothing bad happened recently. Affect Pacifist AI players and Militarist, but on the condition of having as much military as them.

AI will not ask for tribute if they don't perceive themselves as stronger in military.

AI will not move into trying a sneak attack and declare it, if they are not ready to commit into it (A military power comparison later prevented declaring war).

Hostile AI will not go into war if he has already another war and doesn't feel stronger than the target.

Technology progression:

Added some interest to unlock hero and ships part of the marketplace.

Added interest to siege modules, when asked for invasion modules.

Raised the AI interest for non-behemoth hull improvements by 50%.

AI will not research most behemoth technologies if they don't have behemoths, or the behemoth types for the upgrades (IE will not research analyser modules without a science behemoth).

Interest for juggernaut techs drops a bit if there's a big disparity between built & queued juggernauts vs researched juggernaut techs.

The Movement Improvement interests for behemoths, military and exploration fixed and improved.

Loosened the importance score required in order to consider building uniques, instead of the just most important system(usually from 1 to 4 systems as the game advances).

Other:

Ship bound (Volyani) will prefer raze to loot when solving a won ground battle.

Added new generator to allow Cravers to use their sacrifice slave action (Feeding pits) to keep approval and to a much lesser extent, rise system food production.

AI will consider more well developed systems to build their unique buildings. Fixed some analysis related to it.

Reduced the interest in queuing planet terraformation a bit.

Latest touches in AI economics, this time centered on planet colonization.

Reduced some manpower interests to up to 30% less if the system is underdeveloped. There are 6 different generators for manpower, will not be of a big impact but maybe just enough to allow a few low cost industry buildings to kick in.

AI will consider colonizing a new planet when possible population is over 66% instead of 75% of the max, the interest will be very low at that point but much better having that elasticity.

AI will consider planet approval penalty considering a 2/3 filled planet instead of a fully filled planet, when looking for a new planet to colonize.

AI will not boost interest of buildings already in progress if they are infinite.

AI will have a much higher interest in building unique buildings, and higher interest in building wonders.

Refinement on the resources need analysis, to properly take into account infinite buildings.

Extended the functionally of the generator that assigns some interest to build the Academy Embassy, to also give interest to other buildings that AI didn't consider building previously:

- Curiosity Improvements without resource gains: Honeycomb Scope, Vault of Governance...
- Ship Enhancement Improvements: Big Data Shipyards, Cravers Dark ops, Real-time Relays...
- Lumeris Shadow Banking.