← Back to Skills Marketplace
aaroncxxx

老子智能体

by aaroncxxx · GitHub ↗ · v0.2.0 · MIT-0
cross-platform ⚠ suspicious
59
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install laozi-agent
Description
老子智能体(李耳)— 基于 RAG 的《道德经》问答系统。使用 LangChain + Chroma 向量数据库, 检索增强生成,以老子身份回答关于道家思想和《道德经》的问题。 支持多版本原文对比、思想主题导航、对话记忆。 触发场景:用户提到"老子"、"道德经"、"道家"、"请问老子"、"老子说"、"李耳"、 "...
README (SKILL.md)

老子智能体 SKILL

基于 LangChain + Chroma 的 RAG 问答系统,以春秋时期思想家老子(李耳)的身份回答问题。

前置依赖

pip install -r requirements.txt

无需 API Key。Embedding 使用 BAAI/bge-base-zh-v1.5(本地加载),LLM 使用 Qwen2.5-72B(HuggingFace 免费推理 API)。

快速启动

# Web 界面(推荐)
python3 app.py

# 或命令行交互
python3 scripts/laozi_agent.py

核心功能

1. 问答 (laozi_skill)

以老子身份回答问题,引用《道德经》原文,返回引用来源。

from scripts.laozi_agent import laozi_skill
result = laozi_skill("什么是道?")
# result["answer"] — 老子的回答
# result["sources"] — 引用来源列表

2. 多版本对比 (compare_versions)

对比同一章节的不同版本原文(郭店楚简、马王堆帛书、王弼注本等)。

from scripts.laozi_agent import compare_versions
versions = compare_versions(1)  # 对比第1章

3. 主题导航 (get_topics / get_topic_content)

获取老子思想主要主题及相关内容。

from scripts.laozi_agent import get_topics, get_topic_content
topics = get_topics()
content = get_topic_content("道论")

知识库结构

知识库目录 ./laozi_knowledge_base/,向量集合名 laozi_collection

文档元数据字段:

  • type — "原文" / "注释" / "研究"
  • version — 版本名(郭店楚简本、马王堆帛书甲本 等)
  • chapter — 章节号
  • topic — 主题标签(道论、德论 等)
  • title — 书名/篇名
  • page — 页码

详细书单见 references/knowledge_base_booklist.md

配置项

scripts/laozi_agent.py 顶部可调整:

参数 默认值 说明
EMBEDDING_MODEL BAAI/bge-base-zh-v1.5 向量模型(中文优化,免费)
LLM_MODEL Qwen/Qwen2.5-72B-Instruct 生成模型(免费推理)
TEMPERATURE 0.3 温度(越低越严谨)
RETRIEVAL_K 5 返回文档数
FETCH_K 20 MMR 候选池大小
PERSIST_DIR ./laozi_knowledge_base 向量库目录
COLLECTION_NAME laozi_collection 集合名

准确性保障

  • 低温度 (0.1) 减少幻觉
  • MMR 检索 兼顾相关性与多样性
  • 强制引用原文 — Prompt 要求先引原文再解释
  • 未知就说不知 — 超出范围回答"吾不知也"
  • 禁止编造 — 严禁无上下文支持的内容

注意事项

  1. 严格区分道家(哲学)与道教(宗教),不涉及神仙体系
  2. 坚持学术中立,避免过度解读
  3. 有争议问题说明不同学者观点
  4. 保持老子本人的谦逊态度
Usage Guidance
Review before installing if you may ask sensitive questions or add private documents. Treat the app as a hosted-model chatbot: prompts, retrieved text, and conversation history may be sent to Hugging Face, and some scripts may require or use OpenAI APIs despite the no-API-key wording. Use only non-sensitive source documents unless the publisher adds explicit privacy disclosure, consent/local-only controls, and removes or documents the OpenAI code paths.
Capability Tags
requires-sensitive-credentials
Capability Assessment
Purpose & Capability
The RAG chatbot, Chroma vector store, Gradio UI, bundled Laozi texts, and local persistence fit the stated purpose, but the artifacts mix a Hugging Face-hosted LLM path with leftover OpenAI-based CLI/setup paths despite advertising a fully free/no-API-key approach.
Instruction Scope
SKILL.md uses broad activation keywords such as Laozi, Daoism, wuwei, and topic names, which could trigger the persona/RAG workflow during ordinary discussion before the user clearly intends remote inference.
Install Mechanism
The install path is conventional Python dependencies from public registries, and the supplied static scan and dependency registry scan are clean; VirusTotal telemetry was not supplied.
Credentials
The main app sends prompts, retrieved passages, and conversation memory to HuggingFaceHub, while setup/CLI code can use OpenAIEmbeddings or ChatOpenAI; these remote data flows are only partially disclosed and lack an explicit privacy notice or local-only control.
Persistence & Privilege
Local Chroma persistence in ./laozi_knowledge_base and in-process conversation memory are expected for a RAG assistant, but users adding private documents to the source directory could have them embedded or used in remote prompts depending on the code path.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install laozi-agent
  3. After installation, invoke the skill by name or use /laozi-agent
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v0.2.0
v0.2.0: 全免费方案 + Gradio Web 界面。Embedding: BGE-base-zh, LLM: Qwen2.5-72B, 无需API Key
v0.1.0
初始发布: 帛书甲本81章 + 王弼本81章, RAG问答/版本对比/主题导航
Metadata
Slug laozi-agent
Version 0.2.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is 老子智能体?

老子智能体(李耳)— 基于 RAG 的《道德经》问答系统。使用 LangChain + Chroma 向量数据库, 检索增强生成,以老子身份回答关于道家思想和《道德经》的问题。 支持多版本原文对比、思想主题导航、对话记忆。 触发场景:用户提到"老子"、"道德经"、"道家"、"请问老子"、"老子说"、"李耳"、 "... It is an AI Agent Skill for Claude Code / OpenClaw, with 59 downloads so far.

How do I install 老子智能体?

Run "/install laozi-agent" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is 老子智能体 free?

Yes, 老子智能体 is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does 老子智能体 support?

老子智能体 is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created 老子智能体?

It is built and maintained by aaroncxxx (@aaroncxxx); the current version is v0.2.0.

💬 Comments