SPL Pure Core V8.0 —— 一套确定性、语义无关的拟人心理动力学内核。它不理解"赞美"或"背叛"，却能复现情绪惯性、创伤、自尊、睡眠加工与防御机制。
SPL Pure Core V8.0 — a standard anthropomorphic agent template built on the SPL audit engine. Causal anchoring eliminates OOC and memory loss, reproducing emotional inertia, trauma, self-esteem and sleep processing.
核心引擎是"躯体"，提供通用的心理动力；映射器是"滤镜"，把事件翻译成内感受向量；feature 模块是企业/政府可自定义的人格模板。
The core engine is the "body" providing universal psychological dynamics; the mapper is the "filter" translating events into interoceptive vectors; the feature modules are enterprise/government-customizable persona templates.
语义无关的心理动力学，所有人共有。
Semantics-agnostic psychological dynamics shared by all humans.
唯一注入"语义"的一层，决定角色灵魂。
The only layer that injects "semantics", deciding the soul of a character.
WorldModel · BiasEngine · IdentityEngine · ValueEngine · GoalEngine —— 教组织如何为不同场景配置一个拟人 Agent。
WorldModel · BiasEngine · IdentityEngine · ValueEngine · GoalEngine — how an organization configures an anthropomorphic agent for different scenarios.
以下七项并非角色自设，而是所有人类共有的心理架构，已写入核心引擎。
The following seven are not character self-settings but the psychological architecture shared by all humans, written into the core engine.
喜悦 · 愤怒 · 恐惧 · 信任 · 疏离 · 张力 · 愧疚 · 羞耻。各维以自适应粘滞度向动态目标松弛，构成瞬时情绪。
Joy · Anger · Fear · Trust · Detachment · Tension · Guilt · Shame. Each dimension relaxes toward a dynamic target with adaptive viscosity, forming instantaneous emotion.
慢变量背景情绪（愉悦 / 紧张 / 精力），惯性极大（约 1 小时半衰期），持续调制认知评估的增益。
Slow-variable background emotion (Pleasure / Tension / Energy) with huge inertia (~1-hour half-life) continuously modulating the gain of cognitive appraisal.
自尊 [0,1] 随事件缓慢修正；低自尊者把失败内归因、放大威胁感——决定"我值多少"的自我感知。
Self-esteem [0,1] slowly revised by events; low self-esteem internalizes failure and amplifies threat — the self-perception of "how much I'm worth".
sleep(hours) 不仅恢复精力，更在"梦境"中衰减情绪电荷、实现恐惧消退——真正的"睡一觉就好"。
sleep(hours) not only restores energy but also decays emotional charge in "dreams" and achieves fear extinction — a real "sleep it off".
expect() 设定对未来的积极/威胁预期；事件兑现产生惊喜，落空产生失望——让 Agent 为未来而感受。
expect() sets positive/threat expectations about the future; fulfillment brings surprise, miss brings disappointment — letting the agent feel for the future.
信念与行为冲突时产生失调张力，超过阈值后自动通过"合理化"消解——Festinger 经典的自我辩护引擎。
When belief conflicts with behavior, dissonance tension arises; beyond a threshold it auto-resolves via "rationalization" — Festinger's classic self-justification engine.
否认 → 合理化 → 压抑，逐级递进。高压下先否认现实，失效则合理化，再失效则压抑入隐压池等待反弹。
Denial → Rationalization → Repression, escalating. Under high pressure deny reality first; if that fails, rationalize; if that fails, repress into the latent pool awaiting rebound.
同一套核心引擎，换上 SujinMapper，即可演绎一个从低信任到共创、再到背叛后缓慢修复的人物弧光。
The same core engine, swapped with SujinMapper, can play out an arc from low trust to co-creation, then slow repair after betrayal.
信任容量低，对批评敏感、对夸奖保留。预期落空会放大负面情绪，自尊在反复试探中轻微动摇。
Low trust capacity, sensitive to criticism, reserved about praise. Unmet expectations amplify negative emotion; self-esteem wavers slightly through repeated probing.
经过睡眠的梦境加工与预期兑现，信任上升、心境转暖、自尊达到高位，进入高效共创状态。
After dream processing during sleep and expectation fulfillment, trust rises, mood warms, self-esteem peaks — entering efficient co-creation.
背叛触发羞耻维度与自尊崩塌，防御层级启动；认知失调在睡眠中逐步合理化，长期愈合但 max_trust 永久降低。
Betrayal triggers the shame dimension and self-esteem collapse; defense layers activate; cognitive dissonance gradually rationalizes during sleep — long-term healing but max_trust permanently lowered.
纯 Python，零依赖。核心不认识任何事件名，所有语义由外部映射器注入。
Pure Python, zero dependencies. The core knows no event names; all semantics are injected by external mappers.
源于同一套因果锚定内核的五个开源项目：处理器、智能体、治理、叙事和虚拟世界引擎。
Five open-source projects extended from the same causally-anchored core: processor, agent, governance, narrative and virtual-world engines.
面向大规模虚拟世界生态的紧凑异构通用处理器，兼顾高性能计算与低功耗日常调度。
A compact, heterogeneous general-purpose processor for large-scale virtual-world ecosystems, balancing high-performance compute with low-power daily scheduling.
Gitee →基于 SPL 审计引擎构建的标准拟人智能体模板，因果锚定消除 OOC 与记忆丢失。
A standard anthropomorphic agent template built on the SPL audit engine; causal anchoring eliminates OOC and memory loss.
决策前理性审计工具——离线、中立、永不替代决策本身。
A pre-decision rational audit tool — offline, neutral, and never replaces the decision.
Gitee →AIGC 故事生成引擎，保持人物与剧情零逻辑漏洞的连贯创作。
An AIGC story-generation engine that keeps characters and plot coherent with airtight logic.
Gitee →AI 驱动、规则不可篡改的开源虚拟世界引擎（"数字文明方舟"）。
An AI-driven, rule-immutable open-source virtual-world engine (the "Digital Civilization Ark").
Gitee →