Qiaomai Skills

v1.0.5

荞麦饼 Skills 是基于八大维度优化的 OpenClaw 智能体操作系统,支持自然语言交互、智能任务规划和动态知识图谱等功能。

0· 104·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for ruiyongwang/qiaomai-skills.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "Qiaomai Skills" (ruiyongwang/qiaomai-skills) from ClawHub.
Skill page: https://clawhub.ai/ruiyongwang/qiaomai-skills
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install qiaomai-skills

ClawHub CLI

Package manager switcher

npx clawhub@latest install qiaomai-skills
Security Scan
Capability signals
Requires sensitive credentials
These labels describe what authority the skill may exercise. They are separate from suspicious or malicious moderation verdicts.
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (memory, knowledge graph, report generator) align with the included Python modules (octo_memory, dynamic_kg, smart_report, case_search, smart_executor, easy_use). Optional third‑party APIs (OpenAI, Perplexity, Grok) are declared as optional in the metadata and SKILL.md, which fits an enhancement toolkit that can optionally call external services.
Instruction Scope
SKILL.md describes features and optional integrations but does not instruct the agent to read unrelated system files, environment variables, or to send data to unknown endpoints. The runtime instructions stay within the stated domain (memory, KG, report generation).
Install Mechanism
There is no install specification (no arbitrary downloads or installers). The package is delivered as source files and a SKILL.md; that is proportionate for a pure-Python toolkit and reduces supply-chain risk compared with remote downloads.
Credentials
No required environment variables are listed. clawhub.json declares three optional API keys (OPENAI_API_KEY, PERPLEXITY_API_KEY, GROK_API_KEY) marked sensitive and optional — this is proportional for optional semantic/search/code features. Ensure you only set keys you intend to use and scope them with least privilege.
Persistence & Privilege
clawhub.json sets a storage path (~/.qiaomai/), implying local persistence of data (memories/graphs). This is expected for a memory/knowledge tool but you should be aware it will likely store data on the host under that path and review or control that directory.
Assessment
This skill appears coherent with its description, but before installing: (1) Review or sandbox its filesystem writes (the registry metadata declares ~/.qiaomai/ as storage); (2) Only provide optional API keys (OpenAI/Perplexity/Grok) if you trust the integration and prefer using scoped keys; (3) If you need stronger guarantees, inspect the full source for any network I/O or os.environ usage (not observed in the provided fragments) or run the skill in an isolated environment first; (4) Periodically inspect and clean ~/.qiaomai/ if you store sensitive data there.

Like a lobster shell, security has layers — review code before you run it.

latestvk977rknj4eqqfbrhrt1b7cbwxs85cadc
104downloads
0stars
5versions
Updated 5d ago
v1.0.5
MIT-0

荞麦饼 Skills (QiaoMai)

OpenClaw 智能体增强工具包:记忆管理 · 知识图谱 · 报告生成

版本: v1.0.5
状态: 稳定版
兼容: OpenClaw 1.x+


简介

荞麦饼 Skills 是为 OpenClaw 平台设计的智能体能力增强工具包,提供三大核心能力:

  • OctoMemory(八层记忆系统):从瞬时记忆到专家级永久记忆的分层管理
  • DynamicKnowledgeGraph(动态知识图谱):实体关系的智能建立与推理
  • SmartReportGenerator(智能报告生成):多格式文档的自动化生成

核心模块

OctoMemory — 八层记忆系统

将 Agent 的记忆分为 8 个层级,从毫秒级瞬时记忆到永久专家知识,实现跨会话的持续学习与知识积累。

记忆层级

层级类型时效
Layer 8瞬时记忆毫秒级
Layer 7工作记忆秒级
Layer 6短期记忆分钟级
Layer 5会话记忆小时级
Layer 4上下文记忆天级
Layer 3长期记忆月级
Layer 2技能记忆年级
Layer 1专家记忆永久

DynamicKnowledgeGraph — 动态知识图谱

构建和维护实体间的语义关系网络,支持知识推理和关联发现。

支持实体类型:人物、组织、概念、事件、文档

关系类型:属于、相关、因果、时序、引用

核心能力:路径发现、社区检测、影响力传播、增量更新、版本管理

SmartReportGenerator — 智能报告生成

基于结构化数据自动生成多格式专业报告,支持 Markdown、HTML、PDF 等输出格式。


性能指标

指标v1.0.5
启动时间~200ms
记忆检索~5ms
并发处理100 任务
报告生成支持 3 种格式

可选集成

本工具包支持以下可选第三方服务(通过环境变量配置):

  • OpenAI API:增强语义理解能力(可选)
  • Perplexity API:实时信息检索增强(可选)
  • Grok API:代码生成辅助(可选)

所有第三方集成均为可选,不配置时工具包仍可正常运行核心功能。


项目结构

qiaomai-skills/
├── SKILL.md          # 本文件
├── clawhub.json      # 元数据
├── core/             # 核心模块
│   ├── octo_memory.py        # 八层记忆系统
│   ├── dynamic_kg.py         # 动态知识图谱
│   ├── smart_report.py       # 报告生成
│   ├── smart_executor.py     # 智能执行引擎
│   ├── smart_db.py           # 数据存储
│   ├── case_search.py        # 类案检索
│   └── easy_use.py           # 易用性接口
└── data/             # 数据目录

版本历史

  • v1.0.5 (2026-04-23) — 安全加固,清理冗余文件,规范元数据声明
  • v1.0.4 (2026-04-20) — 统一元数据,声明 API keys
  • v1.0.3 (2026-04-18) — 优化显示格式
  • v1.0.2 (2026-04-15) — 简化描述,精简标签
  • v1.0.1 (2026-04-13) — 首次发布更新
  • v1.0.0 (2026-04-12) — 初始发布

许可证

MIT License


荞麦饼 Skills — 营养全面,易于消化,百搭实用。

Comments

Loading comments...