Metadata

name: neural-memory
description: 神经网络启发的记忆系统，支持激活扩散和联想检索。安装后即可使用本地模式，配置 LLM 后可启用智能意图分析。/ Neural network-inspired memory with activation spreading. Works out-of-box in local mode; configure LLM for smart intent analysis.
metadata: {"openclaw":{"requires":{"bins":["python"]},"install":[{"id":"python","kind":"python","label":"Python 3.10+ required / 需要 Python 3.10+"}]}}

Neural Memory / 神经记忆系统

Works out of the box! / 开箱即用！

No configuration required for basic usage. Install and start using immediately. 基础使用无需配置，安装后即可立即使用。

Quick Start / 快速开始

1. Install / 安装

npx clawhub install neural-memory-cn

2. Initialize / 初始化

import sys
sys.path.insert(0, "~/.openclaw/skills/neural-memory-cn/scripts")

from thinking import ThinkingModule

# Auto-detects and creates storage / 自动检测并创建存储
memory = ThinkingModule()

# Start using! / 开始使用！
result = memory.think("民航安全")
print(result)

3. Learn New Knowledge / 学习新知识

memory.learn_and_think(
    content="深度学习是机器学习的一个分支...",
    concept_name="深度学习",
    concept_type="concept",
    tags=["AI", "机器学习"]
)
memory.save()

Usage Modes / 使用模式

Mode	Description	LLM Required
`smart`	Intent analysis + activation spreading / 意图分析 + 激活扩散	Optional
`exact`	Direct neuron lookup / 直接神经元查找	No
`associative`	Hybrid mode / 混合模式	Optional

# Local mode (no LLM needed) / 本地模式（无需 LLM）
result = memory.think("民航安全怎么样", mode="smart")

# Exact lookup / 精确查找
result = memory.think("中医", mode="exact")

Optional: Enable LLM for Better Analysis / 可选：启用 LLM 获得更好分析

English: For enhanced intent understanding, configure an LLM provider.

中文: 要增强意图理解能力，可配置 LLM 提供商。

Method 1: Environment Variables / 方式1：环境变量

export NEURAL_MEMORY_LLM_API_KEY="your-api-key"
export NEURAL_MEMORY_LLM_BASE_URL="https://openrouter.ai/api/v1"
export NEURAL_MEMORY_LLM_MODEL="openai/gpt-3.5-turbo"

Method 2: Setup Script / 方式2：安装脚本

python ~/.openclaw/skills/neural-memory-cn/scripts/setup.py \
    --api-key "your-key" \
    --base-url "https://openrouter.ai/api/v1" \
    --model "openai/gpt-3.5-turbo"

Method 3: Edit Config / 方式3：编辑配置文件

Edit ~/.openclaw/neural-memory/config.yaml:

thinking:
  intent:
    use_llm: true
    llm_api_key: "your-key"
    llm_base_url: "https://openrouter.ai/api/v1"
    llm_model: "openai/gpt-3.5-turbo"

Core Concepts / 核心概念

Concept	Description (EN)	Description (CN)
Neuron	Knowledge unit (concept, fact, experience)	知识单元（概念、事实、经验）
Synapse	Connection between neurons with weight	神经元间的连接，带权重
Activation Spreading	Memory retrieval through association	通过联想进行记忆检索
Intent Layer	Query understanding (optional LLM)	查询理解（可选 LLM）

Define Knowledge Domains / 定义知识领域

Edit domain_hints.json for accurate spreading:

{
  "民航": ["民航安全", "人为因素", "SMS"],
  "中医": ["中医诊断", "中药"],
  "系统思维": ["AnyLogic", "系统动力学"]
}

API Summary / API 摘要

# Query / 查询
result = memory.think("query", mode="smart")

# Learn / 学习
memory.learn_and_think(content, name, type, tags)

# Get stats / 获取统计
stats = memory.get_thinking_stats()

# Save / 保存
memory.save()

File Structure / 文件结构

~/.openclaw/neural-memory/
├── config.yaml              # Configuration / 配置文件
└── memory_long_term/
    ├── neurons.json         # All neurons / 所有神经元
    ├── synapses/            # Connections / 连接
    └── domain_hints.json    # Domain definitions / 领域定义

Documentation / 文档

API Reference: references/api.md
Architecture: references/architecture.md

Troubleshooting / 故障排除

Issue	Solution
Module not found	Check sys.path points to scripts/
Empty results	Add neurons with learn_and_think()
LLM not working	Check API key and base URL in config
Slow queries	Increase hot_cache_size in config

Neural Memory CN

安装技能