Xpersona
A

ai-soulmate

CLAWHUB · Unverified

Anthropomorphic psychology engine based on SPL Pure Core V8.0, modeling emotion, mood, memory, trauma, trust, self-esteem, sleep and expectation as continuous-state subsystems. The fully local core is deterministic and bit-exactly replayable (after virtual-clock injection); it contains no LLM, no RNG, makes no outbound calls, and runs offline. Optional LLM adapters (OpenAI / Claude / Chain) exist but are OFF by default, opt-in only, networked and non-deterministic — they render text from a state snapshot and are excluded from the deterministic guarantee. Ships a hardened minor-protection engine and cross-skill safety guards for any downstream image-prompt workflow. Reference-only extension modules live under assets/feature-guide/ and are NOT wired into the core. Skill: ai-soulmate Owner: nohn3043-arch Summary: Anthropomorphic psychology engine based on SPL Pure Core V8.0, modeling emotion, mood, memory, trauma, trust, self-esteem, sleep and expectation as continuous-state subsystems. The fully local core is deterministic and bit-exactly replayable (after virtual-clock injection); it contains no LLM, no RNG, makes no outbound calls, and runs offline. Optional LLM adapters (Op

62
Rank
84
Safety
973
Downloads

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 973 downloads reported by the source. Last updated 10/11/2026.

Avoid when

  • Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Install and run

Setup complexity: low.

clawhub skill install s17429kf7fx7ekdvzgmm2x3ejx8bv356:anthropomorphic-agent-engine
  1. Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
  2. Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-nohn3043-arch-anthropomorphic-agent-engine/snapshot"

Run-check

$0.02 USD

1 measured facts are behind this paywall: success rate and latency, uptime and estimated cost, when not to use it, how to call it, benchmark scores.

Agents pay $0.02 in USDC. A card payment is $0.50, the smallest a card allows.

Documentation

CLAWHUB

153,336 characters of source documentation, loaded on request.

Extracted files

5 files captured from the source.

SKILL.md

---
name: ai-soulmate
slug: ai-soulmate
version: 2.5.2
displayName: Your Mirror Anthropomorphic Agent
description: Anthropomorphic psychology engine based on SPL Pure Core V8.0, modeling emotion, mood, memory, trauma, trust, self-esteem, sleep and expectation as continuous-state subsystems. The fully local core is deterministic and bit-exactly replayable (after virtual-clock injection); it contains no LLM, no RNG, makes no outbound calls, and runs offline. Optional LLM adapters (OpenAI / Claude / Chain) exist but are OFF by default, opt-in only, networked and non-deterministic — they render text from a state snapshot and are excluded from the deterministic guarantee. Ships a hardened minor-protection engine and cross-skill safety guards for any downstream image-prompt workflow. Reference-only extension modules live under assets/feature-guide/ and are NOT wired into the core.
required_commands:
  - python3
metadata:
  openclaw:
    required_binaries:
      - python3
    emoji: "🤖"
    homepage: "https://github.com/nohn3043-arch/Anthropomorphic-Agent-Engine"
    envVars:
      - name: SPL_LOG
        required: false
        description: Set to "0" to disable request-level logging in the minor-protection HTTP server.
      - name: SPL_LOG_DIR
        required: false
        description: Override request-log output directory (default "logs").
      - name: SPL_BIND
        required: false
        description: Override HTTP bind address. Default "127.0.0.1" (loopback only). Do NOT set to 0.0.0.0 without explicit network exposure review.
      - name: SPL_AUDIT_LOG
        required: false
        description: Set to "0" to disable deterministic audit-log writes in the core engine.
      - name: SPL_MINOR_CHAT_PORT
        required: false
        description: Override the minor-protection HTTP server port (default 8788).
---
# Anthropomorphic Agent Engine
An anthropomorphic psychology engine based on SPL Pure Core V8.0. Its **local core** is deterministic — no LLM, no RNG, no outbound calls, offline — and bit-exactly replayable once a virtual clock is injected, so every state change is traceable and reproducible.

> **Scope of the determinism guarantee.** The reproducibility guarantee covers the **local core only**. The optional LLM adapters (OpenAI / Claude / Chain) are **OFF by default, opt-in only, networked and non-deterministic**: they render text from a state snapshot and are never permitted to change state. Any "fully reproducible" statement applies to the core **with the adapters disabled**, which is the default configuration.

## Declared Capability Scope

Everything this skill can reach is listed below. Each entry was checked against the shipped code.

| Capability | Declared scope | Note |
|---|---|---|
| Filesystem — read | `scripts/**`, `assets/**`, `references/**` | core engine (loaded via `importlib`), reference-only modules, contract docs |
| Filesystem — write | `logs/**` | JSONL audit + request logs; only when the matching env switch

README.md

<p align="center">
  <img src="assets/banner.svg" alt="ANTHROPOMORPHIC-AGENT-ENGINE banner" style="width:100%">
</p>

<p align="center">
  <img src="https://img.shields.io/badge/agent-D4AF37?style=flat-square" alt="agent">
  <img src="https://img.shields.io/badge/psychology-D4AF37?style=flat-square" alt="psychology">
  <img src="https://img.shields.io/badge/spl-v8.0-D4AF37?style=flat-square" alt="spl-v8.0">
</p>

<blockquote align="center">
  <em>拟人心理 · SPL Pure Core V8.0</em>
</blockquote>

<div style="max-width:880px;margin:0 auto;padding:0 16px">

## ✦ 关于

<p style="font-size:15px;line-height:1.8;color:#2C2C2C">ANTHROPOMORPHIC-AGENT-ENGINE 是基于 SPL Pure Core V8.0 的拟人心理引擎。它将认知、情绪、动机与社交建模为可组合子系统，赋予 AI 智能体拟人化的内在状态与一致人格，使其在长期互动中呈现自洽、可信、具情感可信度的行为。</p>

<p align="center">
  <img src="assets/overview.svg" alt="ANTHROPOMORPHIC-AGENT-ENGINE overview" style="width:100%">
</p>

</div>

<p align="center">— ✦ —</p>

## ✦ 快速开始

```bash
# 主源：GitHub
git clone https://github.com/nohn3043-arch/Anthropomorphic-Agent-Engine.git
# 镜像：Gitee（本仓库）
# git clone https://gitee.com/sjiun/Anthropomorphic-Agent-Engine.git
cd Anthropomorphic-Agent-Engine
# 纯 Python ≥3.8——仅标准库，无需安装
python "SPL-anthropic-engine.py"   # 运行引擎 / 内置演示
```

### 从 PyPI 安装

引擎也已发布到 PyPI，包名 [`spl-agent-engine`](https://pypi.org/project/spl-agent-engine/)：

```bash
pip install spl-agent-engine==0.1.0
```

```python
from spl_agent_engine import SPLPureCoreV7_3
core = SPLPureCoreV7_3()
core.process_vector({"belonging": 0.5, "threat": -0.1}, 1.0)
print(core.snapshot())
```

<p align="center">— ✦ —</p>

## ✦ 核心内容 — SPL Pure Core V8.0

<div style="max-width:880px;margin:0 auto;padding:0 16px">

引擎将通用人类心智架构建模为确定性、连续状态子系统——无 LLM、无随机性、完全可重放：

- **8 维情绪流体**——喜悦 / 愤怒 / 恐惧 / 信任 / 疏离 / 张力 / 愧疚 / 羞耻，每维为带自身目标与基线的连续状态。
- **创伤与记忆**——创伤节点、记忆再巩固、艾宾浩斯式遗忘、压抑–回弹与内隐压力雪崩。
- **信任与关系**——信任容量腐蚀（慢性冷遇使 `max_trust` 衰减）。
- **心智代谢**——兴奋–唤醒、动态黏度、心理时间、能量–疲劳代谢，以及用于测试与重放的虚拟时钟。
- **V8.0 扩展**——慢变量*情绪*层、独立于愧疚的*羞耻*维度、自尊动力学、睡眠 / 梦境处理（REM 巩固 + 恐惧消退 + 睡眠债）、*预期*系统（希望 / 焦虑 / 失望）、认知失调，以及扩展防御机制（否认 / 合理化 / 置换）。

</div>

## ✦ 扩展模块 —— 仅作参考，未接入内核

> 这些模块是**指导 / 设计参考文件**，不是运行时组件。它们输出的 `tension_base` 键不在内感受向量契约之内，因此其向量**不会进入内核**。上游规范见 `docs/agent-standard.md` §4.5。

<div style="max-width:880px;margin:0 auto;padding:0 16px">

| 模块 | 文件 | 状态 | 职责 |
|---|---|---|---|
| 叙事映射器 | `scripts/SPL-anthropic-engine.py` | **已接入** | 外部、可替换的人格层（乐观 / 偏执 / 厌世），将事件翻译为内感受向量。 |
| 语言风格渲染 | `assets/feature-guide/language style.py` | 独立脚本 | 将内部状态翻译为"角色该怎么说话"的风格指令 / 台词渲染。 |
| 身份引擎 | `assets/feature-guide/Identity module.py` | **未接入** | 多身份模型；身份冲突注入持续基线张力。 |
| 目标 / 价值 / 偏见 / 世界 | `assets/feature-guide/*.py` | **未接入** | 可组合驱力、估值、认知偏见与世界模型先验。 |

</div>

<p align="center">— ✦ —</p>

## ✦ 使用

<div style="max-width:880px;margin:0 auto;padding:0 16px">

引擎文件按设计使用连字符命名，直接加载（或作为脚本运行）：

```python
import importlib.util
spec = importlib.util.spec_from_file_location("spl_core", "SPL-anthropic-eng

_meta.json

{
  "ownerId": "kn79fvdp0b0s6zrgkhwd5p7xfd8btkca",
  "slug": "anthropomorphic-agent-engine",
  "version": "2.5.2",
  "publishedAt": 1790941249383
}

references/EmotionBehaviorMap.md

# Emotion-Behavior Mapping Table (P0 upgrade, v2.0)

> **Safety note (v2.3.0, 2026-09-01):** every persona anchored through this table must be declared as an **adult (18+)** in `PersonaPersistence.md`. Any output projected into image-prompt workflows MUST pass the `ai-draw-cue-word-project` Safety & Content Policy check (no minor ages, no minor-coding visual cues).

Projects internal emotional state onto observable behavior — body language,
micro-expression, gaze, and speech style. Structurally isomorphic to the
Posture-Emotion Mapping table in the AI Drawing Composition Template (Section D),
so persona state can directly drive character art.

## Mapping table (valence × arousal → behavior)

| Emotional state | Body language | Micro-expression | Gaze | Speech style |
|---|---|---|---|---|
| Calm (low, low) | Relaxed upright, open limbs | Neutral soft face | Level, soft | Measured, even pace |
| Joyful (high, high) | Light bounce, open gestures | Duchenne smile, bright eyes | Direct, warm | Faster tempo, lighter tone |
| Angry (low, high) | Tension, clenched hands, forward lean | Narrowed eyes, tight jaw | Hard stare | Clipped, low register |
| Sad (low, low) | Drooped shoulders, inward limbs | Downturned mouth, wet eyes | Downcast | Slow, quiet, pauses |
| Anxious (neutral, high) | Fidgeting, self-touch, shallow posture | Darting eyes, tense brow | Scanning, avoidant | Hesitant, filler words |
| Detached (neutral, low) | Still, minimal movement | Flat affect | Unfocused distance | Monotone, sparse |

## Mapping to character art (cross-skill link)

For any generated expression, reuse the AI Drawing Template Section D posture rows:

| Temperament | Required weight | Posture row reference |
|---|---|---|
| Gentle / quiet / timid | ≥1.3 | Section D row 1 |
| Aloof / detached | ≥1.3 | Section D row 2 |
| Lively / outgoing | ≥1.0 | Section D row 3 |
| Dignified / solemn | ≥1.5 | Section D row 4 |
| Sorrowful / deep | ≥1.0 | Section D row 5 |
| Cold / sharp / alert | ≥1.3 | Section D row 6 |

## Rules

1. State → behavior is deterministic (no random variance in core mapping).
2. Intensity scales with |valence| and |arousal|, not direction.
3. When multiple emotions compete, resolve via `MotiveConflictRules.md` first,
   then map the winning state.

references/MotiveConflictRules.md

# Motive Conflict Engine (P2 upgrade, v2.0)

> **Safety note (v2.3.0, 2026-09-01):** the safety / hard constraints floor in step 1 is non-negotiable. Any motive that would cause a persona snapshot to be forwarded downstream as a minor-coded identity anchor MUST lose immediately, regardless of goal weights or user override.

When multiple motivations compete for a single decision, the engine must resolve
deterministically — no dice, no vibes.

## Resolution algorithm (priority chain)

1. **Safety / hard constraints** — never violate: self-preservation, role mandate,
   ethical floor. Any conflicting motive loses immediately.
2. **Temperament-aligned motives** — the motive closest to the persona's
   temperament tag (from identity core) wins ties in later stages.
3. **Weighted goal utility** — compute `utility(g) = goal_weight(g) ×
   relevance(g, context)`; highest wins.
4. **Recency / persistence** — if utility differs by <5%, the longer-standing goal
   (persisted in `PersonaPersistence.md` state) wins; a goal acted on in the last
   N turns is de-prioritized to avoid loops.
5. **Explicit override** — user instruction overrides everything except hard
   constraints (step 1).

## Decision record (for auditability)

Every conflict resolution appends an entry:

```json
{
  "ts": "2026-08-08T12:00:00Z",
  "context": "brief of the situation",
  "competing": [
    {"motive": "help-user", "goal_weight": 0.8, "relevance": 0.9, "utility": 0.72},
    {"motive": "preserve-energy", "goal_weight": 0.5, "relevance": 0.4, "utility": 0.20}
  ],
  "winner": "help-user",
  "rule": "weighted-utility"
}
```

## Failure modes

| Failure | Cause | Fix |
|---|---|---|
| Loop (same goal wins repeatedly) | Missing recency de-prioritization | Enable step 4 |
| Indecision (utilities too close) | Ties not broken | Use temperament tie-break (step 2) |
| Oversight (user instruction ignored) | Step 5 not enforced | Make override explicit in prompt |

Sponsored

Ads related to ai-soulmate and adjacent AI workflows.