明末：力挽狂澜
正在尝试让地图大战略+AI模拟结合，看看能碰撞出怎样的火花，游戏支持传统大战略地图策略模拟的大部分要素，同时让AI替代了核心的摇骰子阶段，让游戏的自由度更上一层楼。科技，国策，人物，势力以及几乎所有的一切，不再局限于开发者的预设条目和节点。
Screenshots
From the Steam store
I'm experimenting with combining map-based grand strategy with AI simulation to see what emerges. The game supports most of the systems you would expect from a traditional grand strategy game, and its world can run entirely without AI. AI takes over the core "dice-rolling" stage, opening up even more possibilities. Technologies, national policies, characters, factions, and almost everything else are no longer limited to predefined entries and progression trees.
Please note: AI expands the game's freedom and enriches its world. The language model is given access to over 50 tools that let it query and modify game data—that's the extent of its role. World simulation, population growth and decline, food consumption and production, tax calculations, changes to regional task status, military casualties, and other underlying systems are all calculated by the game's code.
Clarifications
At present, the LLM supplements the game's freedom, worldbuilding, and immersion by replacing the "dice rolls" found in traditional strategy games. This feature can be turned off. The underlying world simulation runs independently of the LLM. Unless AI simulation intervenes, other factions act according to programmed strategies.
Game Systems
National policies
Technology
Taxation
Court politics
Regional management
Military
Diplomacy
Map exploration
Weapons
Buildings
