Pick AI Skills in 1 Minute
Tested AI skills and workflows, worth your minute.
Pre-install Skill Security Vetting
2026-03-21
skill-vetter vs skill-auditor
You need to install a new skillk, but one bad install can become a supply-chain risk.
This review focuses on one practical decision: which pre-install security gate should you run first?
Conclusion:
Use skill-vetter as your default install gate. Use skill-auditor for second-pass review on high-risk candidates.
Core difference:
skill-vetter is optimized for fast pre-install gate decisions; skill-auditor is optimized for evidence-driven 6-step review.
skill-vetter is good at:
- pre-install permission red-flag scan
- quick go/no-go install decision
- structured verdict output for team approval
skill-auditor is good at:
- 6-step evidence-driven audit
- second-pass review for high-risk candidates
- audit evidence capture for retrospective rule tuning
Popularity metrics:
- skill-vetter: GitHub stars: 34; Installs: 9.3K/week
- skill-auditor: GitHub stars: 34; Installs: 207/week
Sources:
- https://skills.sh/useai-pro/openclaw-skills-security/skill-vetter
- https://skills.sh/useai-pro/openclaw-skills-security/skill-auditor
- https://github.com/UseAI-pro/openclaw-skills-security
Efficiency without a security gate is just delayed risk.
You need to install a new skillk, but one bad install can become a supply-chain risk.
This review focuses on one practical decision: which pre-install security gate should you run first?
Conclusion:
Use skill-vetter as your default install gate. Use skill-auditor for second-pass review on high-risk candidates.
Core difference:
skill-vetter is optimized for fast pre-install gate decisions; skill-auditor is optimized for evidence-driven 6-step review.
skill-vetter is good at:
- pre-install permission red-flag scan
- quick go/no-go install decision
- structured verdict output for team approval
skill-auditor is good at:
- 6-step evidence-driven audit
- second-pass review for high-risk candidates
- audit evidence capture for retrospective rule tuning
Popularity metrics:
- skill-vetter: GitHub stars: 34; Installs: 9.3K/week
- skill-auditor: GitHub stars: 34; Installs: 207/week
Sources:
- https://skills.sh/useai-pro/openclaw-skills-security/skill-vetter
- https://skills.sh/useai-pro/openclaw-skills-security/skill-auditor
- https://github.com/UseAI-pro/openclaw-skills-security
Efficiency without a security gate is just delayed risk.
skill-vetter vs skill-auditor
当你要安装一个新 skill 时，一次错误安装就可能演变为供应链风险。
这篇评测聚焦一个实际决策：安装前你应先跑哪一道安全闸门？
结论：
把 skill-vetter 作为默认安装闸门；对高风险候选项，再用 skill-auditor 做第二轮复核。
核心差异：
skill-vetter 面向“快速安装前闸门决策”；skill-auditor 面向“基于证据的 6 步审查”。
skill-vetter 的优势：
- 安装前权限红旗扫描
- 快速给出 go/no-go 安装决策
- 输出结构化结论，便于团队审批
skill-auditor 的优势：
- 基于证据的 6 步审计流程
- 面向高风险候选项的第二轮审查
- 形成审计证据，便于回溯调优规则
热度指标：
- skill-vetter: GitHub stars: 34; Installs: 9.3K/week
- skill-auditor: GitHub stars: 34; Installs: 207/week
来源：
- https://skills.sh/useai-pro/openclaw-skills-security/skill-vetter
- https://skills.sh/useai-pro/openclaw-skills-security/skill-auditor
- https://github.com/UseAI-pro/openclaw-skills-security
没有安全闸门的效率，只是风险的延迟兑现。
当你要安装一个新 skill 时，一次错误安装就可能演变为供应链风险。
这篇评测聚焦一个实际决策：安装前你应先跑哪一道安全闸门？
结论：
把 skill-vetter 作为默认安装闸门；对高风险候选项，再用 skill-auditor 做第二轮复核。
核心差异：
skill-vetter 面向“快速安装前闸门决策”；skill-auditor 面向“基于证据的 6 步审查”。
skill-vetter 的优势：
- 安装前权限红旗扫描
- 快速给出 go/no-go 安装决策
- 输出结构化结论，便于团队审批
skill-auditor 的优势：
- 基于证据的 6 步审计流程
- 面向高风险候选项的第二轮审查
- 形成审计证据，便于回溯调优规则
热度指标：
- skill-vetter: GitHub stars: 34; Installs: 9.3K/week
- skill-auditor: GitHub stars: 34; Installs: 207/week
来源：
- https://skills.sh/useai-pro/openclaw-skills-security/skill-vetter
- https://skills.sh/useai-pro/openclaw-skills-security/skill-auditor
- https://github.com/UseAI-pro/openclaw-skills-security
没有安全闸门的效率，只是风险的延迟兑现。
Browser Automation
2026-03-19
Browser automation AI skills
browser-use vs agent-browser
Core difference
browser-use is built for fast first delivery.
agent-browser is built for controlled, extensible workflows.
Where browser-use wins:
Fast form and page loops
Low-friction validation
Short-cycle demo delivery
Where agent-browser wins:
Controlled multi-step flows (especially after login)
Browser-agent extensibility
Long-cycle workflow integration
Safety
browser-use: Medium
Browser automation with real page actions. Avoid sensitive sessions.
agent-browser: Medium
More controllable execution. Apply least-privilege access.
Popularity metrics
browser-use: GitHub stars 80,869 | Weekly installs 49.4K
agent-browser: GitHub stars 22,410 | Weekly installs 100.4K
Data checked at: 2026-03-16 01:49 UTC
Sources: GitHub API, skills.sh
Quick read
browser-use: Fast first run, low setup friction, strong for quick demos
agent-browser: Higher control, better extensibility, built for production workflows
Skill links
browser-use: https://github.com/browser-use/browser-use
agent-browser: https://github.com/vercel-labs/agent-browser
browser-use vs agent-browser
Core difference
browser-use is built for fast first delivery.
agent-browser is built for controlled, extensible workflows.
Where browser-use wins:
Fast form and page loops
Low-friction validation
Short-cycle demo delivery
Where agent-browser wins:
Controlled multi-step flows (especially after login)
Browser-agent extensibility
Long-cycle workflow integration
Safety
browser-use: Medium
Browser automation with real page actions. Avoid sensitive sessions.
agent-browser: Medium
More controllable execution. Apply least-privilege access.
Popularity metrics
browser-use: GitHub stars 80,869 | Weekly installs 49.4K
agent-browser: GitHub stars 22,410 | Weekly installs 100.4K
Data checked at: 2026-03-16 01:49 UTC
Sources: GitHub API, skills.sh
Quick read
browser-use: Fast first run, low setup friction, strong for quick demos
agent-browser: Higher control, better extensibility, built for production workflows
Skill links
browser-use: https://github.com/browser-use/browser-use
agent-browser: https://github.com/vercel-labs/agent-browser
Browser automation AI skills
browser-use vs agent-browser
核心差异
browser-use 更适合快速首版交付。
agent-browser 更适合可控、可扩展的工作流。
browser-use 的优势：
快速表单与页面循环
低摩擦验证
短周期 Demo 交付
agent-browser 的优势：
可控的多步骤流程（尤其登录后流程）
更强的 browser-agent 扩展能力
更适合长周期工作流集成
安全性
browser-use：中等
具备真实页面动作能力，避免在敏感会话中直接使用。
agent-browser：中等
执行过程更可控，建议按最小权限原则配置。
热度指标
browser-use：GitHub stars 80,869 | Weekly installs 49.4K
agent-browser：GitHub stars 22,410 | Weekly installs 100.4K
数据时间：2026-03-16 01:49 UTC
数据来源：GitHub API, skills.sh
快速结论
browser-use：上手快、配置轻，适合快速验证与演示
agent-browser：控制力与扩展性更强，更适合生产工作流
技能链接
browser-use: https://github.com/browser-use/browser-use
agent-browser: https://github.com/vercel-labs/agent-browser
browser-use vs agent-browser
核心差异
browser-use 更适合快速首版交付。
agent-browser 更适合可控、可扩展的工作流。
browser-use 的优势：
快速表单与页面循环
低摩擦验证
短周期 Demo 交付
agent-browser 的优势：
可控的多步骤流程（尤其登录后流程）
更强的 browser-agent 扩展能力
更适合长周期工作流集成
安全性
browser-use：中等
具备真实页面动作能力，避免在敏感会话中直接使用。
agent-browser：中等
执行过程更可控，建议按最小权限原则配置。
热度指标
browser-use：GitHub stars 80,869 | Weekly installs 49.4K
agent-browser：GitHub stars 22,410 | Weekly installs 100.4K
数据时间：2026-03-16 01:49 UTC
数据来源：GitHub API, skills.sh
快速结论
browser-use：上手快、配置轻，适合快速验证与演示
agent-browser：控制力与扩展性更强，更适合生产工作流
技能链接
browser-use: https://github.com/browser-use/browser-use
agent-browser: https://github.com/vercel-labs/agent-browser
Impeccable
2026-03-18
Impeccable is an enhanced frontend-design skill pack designed to give AI tools a clearer understanding of design.
Instead of relying on raw generation, it introduces a structured workflow:
/audit to identify issues
/arrange to refine layout and spacing
/typeset to improve typography and hierarchy
/polish to elevate overall quality
/teach-impeccable to align outputs with a consistent design system
By embedding design language and review logic directly into AI workflows, Impeccable helps transform outputs from acceptable to well-structured and production-ready.
https://impeccable.style/
Hype is easy. Useful tools are rare.
Instead of relying on raw generation, it introduces a structured workflow:
/audit to identify issues
/arrange to refine layout and spacing
/typeset to improve typography and hierarchy
/polish to elevate overall quality
/teach-impeccable to align outputs with a consistent design system
By embedding design language and review logic directly into AI workflows, Impeccable helps transform outputs from acceptable to well-structured and production-ready.
https://impeccable.style/
Hype is easy. Useful tools are rare.
Impeccable 是一个增强版的 frontend-design skill pack，帮助 AI 工具更准确地理解设计。
它不是只依赖原始生成，而是引入结构化工作流：
/audit：识别问题
/arrange：优化布局与间距
/typeset：优化排版与信息层级
/polish：提升整体完成度
/teach-impeccable：让输出对齐统一设计系统
通过把设计语言与审查逻辑直接嵌入 AI 工作流，Impeccable 可以把“可用”输出提升为“结构清晰、可落地”的产出。
https://impeccable.style/
制造热度很容易，真正有用的工具很少。
它不是只依赖原始生成，而是引入结构化工作流：
/audit：识别问题
/arrange：优化布局与间距
/typeset：优化排版与信息层级
/polish：提升整体完成度
/teach-impeccable：让输出对齐统一设计系统
通过把设计语言与审查逻辑直接嵌入 AI 工作流，Impeccable 可以把“可用”输出提升为“结构清晰、可落地”的产出。
https://impeccable.style/
制造热度很容易，真正有用的工具很少。