by 安全扫描
OpenClaw
可疑
medium confidenceThe skill's code and instructions match its stated purpose (a local rule-based intent parser with optional local Ollama enhancement), but the README encourages running a remote install script (curl | bash) and model downloads which are disproportionate to a simple instruction-only skill and raise installation/exfiltration risk.
评估建议
The skill itself is coherent: it parses intents with a local rule engine and optionally uses a local Ollama model at localhost:11434. However, do NOT run the README's curl | bash installer without inspecting it first — that single command fetches and executes a remote script and is the primary risk here. If you want to use the Ollama enhancement, prefer installing Ollama and pulling models through verified package managers (brew, official Ollama docs) and inspect any model sources. Recommended s...详细分析 ▾
✓ 用途与能力
Name/description (intent parser) match the included code: a rule-based transformer with an optional call to a local Ollama model. No unrelated credentials, binaries, or paths are requested.
✓ 指令范围
SKILL.md only instructs using the rule engine and optionally installing/using Ollama locally. The runtime instructions and code operate on the provided input and call only localhost (Ollama) if available; they do not read arbitrary files or request unrelated secrets.
⚠ 安装机制
The skill is instruction-only (no official install spec), but README includes a high-risk curl -sL https://raw.githubusercontent.com/.../install.sh | bash one-liner. Running an unknown remote install script is dangerous and not necessary for the rule-engine behavior; the optional Ollama commands (brew install, ollama pull) are more reasonable but do involve downloading models from external sources.
✓ 凭证需求
No environment variables, credentials, or config paths are required by the skill. The code only targets a localhost Ollama endpoint and does not access system secrets.
✓ 持久化与权限
Skill does not request permanent presence (always:false) and does not modify other skills or system configs. Autonomous invocation is allowed (platform default) but not combined with other high privileges.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv2.2.02026/3/26
Updated README, merged clarity-engine into clarity-ai
● 可疑
安装命令 点击复制
官方npx clawhub@latest install clarity-ai
镜像加速npx clawhub@latest install clarity-ai --registry https://cn.clawhub-mirror.com
技能文档
Transform messy input into crystal-clear instructions
What It Does
| Messy Input | Clean Output |
|---|---|
| "Hi, about Python for loops, why is it running so slow?" | Intent: Performance Language: Python Goal: Analyze slow loop |
Features
| Feature | Description |
|---|---|
| 🎯 7 Intent Types | Debug, Performance, Create, Explain, Modify, Review, Learn |
| ⚡ Millisecond | Local rule engine, no waiting |
| 🔒 Privacy | 100% local processing |
| 🌏 Bilingual | Chinese & English |
Supported Intents
| Intent | Keywords |
|---|---|
| Code Debug | debug, error, bug, issue |
| Performance | slow, performance, optimize |
| Code Creation | write, create, generate |
| Concept Explain | what is, explain, principle |
| Code Modify | modify, refactor |
| Code Review | review, check |
| Learning | teach, learn |
Usage
Activate: "开启精准模式" or "Enable precision mode"
Input: "Can you help me check this Python code?"
Output:
## 🎯 Precision Instruction📌 Intent: Code Debug
💻 Language: Python
🎯 Goal: Find and fix code issues
Deactivate: "关闭精准模式" or "Disable precision mode"
Installation
clawhub install clarity-ai
Optional: Ollama Enhancement
brew install ollama
ollama pull qwen:0.5b
brew services start ollama
Technical Details
- Speed: Millisecond response (rule engine)
- Privacy: 100% local
- Languages: 20+ programming languages
- Intelligence: 7 intent types
Author
Boss - Built for better AI communication
数据来源：ClawHub ↗ · 中文优化：龙虾技能库
OpenClaw 技能定制 / 插件定制 / 私有工作流定制
免费技能或插件可能存在安全风险，如需更匹配、更安全的方案，建议联系付费定制