技能详情（站内镜像，无评论）
作者：Azrael @azraelji
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 0 · 194 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：良性 · OpenClaw ：良性
Package：azraelji/testhy
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：良性
OpenClaw 评估
This is an instruction-only architecture manual for a Claude Skills framework; it does not request credentials, install software, or instruct the agent to access unrelated system data.
目的
The name/description say it's a sample/framework for building Skills and the SKILL.md is an architecture guide. There are no declared env vars, binaries, or config paths that would be inconsistent with a documentation/manual artifact.
说明范围
SKILL.md contains descriptive documentation (architecture, module names and file examples). It does not instruct runtime actions, file reads, credential access, or network exfiltration. References to filenames (main.py, core/*.py) are explanatory, not commands to access the host.
安装机制
There is no install spec and no code files to write or execute. Instruction-only skills are lowest-risk because nothing is downloaded or executed by the installer.
证书
No environment variables, credentials, or config paths are requested. Nothing in the content implies a need for unrelated secrets or elevated access.
持久
Skill flags are default (not always), and autonomous invocation remains allowed (platform default). The skill does not request persistent presence or modify other skills/configuration in its documentation.
综合结论
This skill is a documentation/manual for a Skills framework and appears safe: it doesn't ask for keys or install code. It also does not provide runnable functionality by itself — installing it will not add tools. If you expect a working Skill, wait for a version that includes code or a trusted source link. Before installing any future updates, re-check for added install scripts, external download URLs, or required environment variables (those …
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「testhy」。简介：testhy is a sample Skill built on the Claude Skills framework enabling natural …。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/azraelji/testhy/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
# Claude Skills 开发框架架构指导手册
## 1. 项目概述 (Overview)
本框架基于 Anthropic Claude 的 Tool Use (Skills) 特性构建，旨在提供一个通过自然语言聊天即可调用各类底层工具的开发脚手架。
框架遵循高内聚、低耦合的设计原则，使开发者能够专注于“技能（Skill）”本身的编写，而无需过度关心 LLM 对话上下文、工具发现与函数执行的复杂逻辑。
## 2. 架构设计 (Architecture Design)
整个系统分为以下几个分层架构：
- **交互层 (CLI/API Interface)**: 负责接收用户输入，展示 Agent 的回复或执行结果。`main.py`
- **代理层 (Agent Layer)**: 负责维护对话历史、与 Claude API 交互，解析模型返回的 Tool Use 请求。`core/agent.py`
- **注册与调度层 (Registry & Dispatcher)**: 维护系统中所有可用的 Skills（工具），向 LLM 声明这些工具的 JSON Schema，并根据 LLM 的指令找到对应的 Python 函数执行器。`core/registry.py`
- **技能层 (Skills Layer)**: 具体的业务代码抽象类及实现，继承自统一的 BaseSkill。`core/skill_base.py`, `skills/*`
## 3. 核心模块说明 (Core Modules)