openclaw 网盘下载
OpenClaw

技能详情(站内镜像,无评论)

首页 > 技能库 > StyleBuddy

OpenClaw首个生活场景Skill - 你的AI穿搭闺蜜。拍照录入衣橱、智能推荐搭配、逛街种草咨询,还有衣橱分析帮你科学购物。支持多模态识别,无需关键词,像跟闺蜜聊天一样自然。

AI 与大模型

许可证:MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本:v0.4.4

统计:⭐ 0 · 177 · 1 current installs · 1 all-time installs

0

安装量(当前) 1

🛡 VirusTotal :良性 · OpenClaw :良性

Package:89kpjddmtb-ui/stylebuddy

安全扫描(ClawHub)

  • VirusTotal :良性
  • OpenClaw :良性

OpenClaw 评估

StyleBuddy's code, docs, and runtime instructions are consistent with an AI wardrobe/dressing assistant; nothing in the provided material demands unrelated credentials or unexpected system access, but review of the remaining omitted files and consideration of privacy (photo storage) is recommended.

目的

Name/description (AI wardrobe, image-based outfit recommendations) match the included code, data assets, and config. Optional integrations (weather, image search, image generation) are declared in config.yaml and make sense for the stated features. There are no declared environment variables or unrelated credentials requested.

说明范围

SKILL.md describes only wardrobe, recommendation, and optional external APIs. The runtime code snippets shown operate on local DB, local asset files, and (optionally) well-scoped external services (Open-Meteo, Serper, SiliconFlow) — consistent with the described capabilities. SKILL.md does not instruct the agent to read unrelated system files or exfiltrate data.

安装机制

There is no install specification (instruction-only metadata), which is low-risk. However the bundle includes Python code and a requirements.txt: running the skill will require installing those Python deps and executing the Python code in the agent environment. No downloads from untrusted URLs or arbitrary installers were presented.

证书

The skill declares no required environment variables or credentials. config.yaml exposes optional API key fields for image search/generation (user-provided) — this is proportional and expected. No evidence that the skill requires unrelated secrets or cloud credentials.

持久

Skill is not always-enabled and uses normal autonomous invocation. It stores user data locally (SQLite and asset image folders) which is expected for this kind of skill. There is no indication it attempts to alter other skills or system-wide configs.

综合结论

This skill appears to be what it says: a local wardrobe manager and outfit recommender. Things to consider before installing: - Privacy: the skill stores photos and metadata in a local SQLite DB and assets/images; avoid uploading very sensitive photos and back up/secure the database if you care about privacy. - Optional APIs: image search/generation providers (Serper, SiliconFlow) require API keys you would add to config.yaml; only provide key…

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「StyleBuddy」。简介:OpenClaw首个生活场景Skill - 你的AI穿搭闺蜜。拍照录入衣橱、智能推荐搭配、逛街种草咨询,还有衣橱分析帮你科学购物。支持多模态识别,无需关键词,…。
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/89kpjddmtb-ui/stylebuddy/SKILL.md
(来源:yingzhi8.cn 技能库)

SKILL.md

打开原始 SKILL.md(GitHub raw)

---
name: stylebuddy
description: OpenClaw首个生活场景Skill - 你的AI穿搭闺蜜。拍照录入衣橱、智能推荐搭配、逛街种草咨询,还有衣橱分析帮你科学购物。支持多模态识别,无需关键词,像跟闺蜜聊天一样自然。
---

# StyleBuddy - AI穿搭闺蜜 🌸

> OpenClaw 生态首个生活场景 Skill

你的专属 AI 穿搭顾问,像闺蜜一样懂你的风格。无论是日常穿搭、衣橱管理还是逛街种草,一句话搞定。

## ✨ 核心能力

### 1. 衣橱智能管理
📸 **拍照录入** — 拍下衣服自动识别类型、颜色、季节  
📂 **自动分类** — 上衣/下装/外套/鞋子智能归档  
👁️ **视觉识别** — 支持多模态模型识别,降级时本地分析,最差也能手动录入

### 2. 每日穿搭推荐
🌤️ **场景感知** — "今天约会穿什么?""明天有雨怎么搭?"  
🎨 **风格匹配** — 根据你的用户画像(身材/风格/偏好)智能推荐  
👗 **成套搭配** — 上衣+下装+鞋子+配饰一键生成

### 3. 逛街种草咨询
📱 **实时参谋** — 看中一件衣服?拍照发给 StyleBuddy  
💡 **搭配建议** — "这件和我衣橱里的米色风衣能搭吗?"  
🛍️ **购买决策** — 基于现有衣橱,给出买或不买的建议

### 4. 衣橱分析洞察
📊 **配置分析** — 了解你的衣橱构成(颜色分布、季节占比、风格偏向)  
🎯 **购物指南** — "缺一件黑色基础款打底衫""夏季单品太多,建议补充秋季"  
💰 **理性消费** — 避免重复购买,优化衣橱性价比

## 🚀 使用示例

像聊天一样自然,无需记指令:

```
今天穿什么?                    → 根据天气和场合推荐搭配
录入这件衬衫                   → 拍照/上传图片,自动识别录入
这件衣服适合我吗?            → 种草咨询,给出搭配建议
帮我分析衣橱                   → 生成衣橱报告和购物建议
明天要见客户,帮我搭配一下    → 场景化推荐
```

## 🏗️ 技术亮点

### 架构设计
- **OpenClaw 原生集成** — 语义理解为主,关键词辅助路由,无需记忆复杂指令
- **智能降级策略** — 多模态识别 → 本地分析 → 手动输入,三级优雅降级
- **用户画像系统** — 记录身材数据、风格偏好,越用越懂你

### 兼容性
- ✅ 支持 Kimi、Claude、GPT 等多模型基座
- ✅ 自动检测视觉能力,无 Vision 也能用
- ✅ 纯文本模型通过本地 PIL 分析图片基础信息

## 📦 版本信息

**v0.4.3** (2026-03-17) — 安全合规版本
- 更新 SKILL.md 描述,准确说明语义+关键词混合路由策略
- 移除开发脚本 (download_images.sh),避免安全误报
- 生成脚本改为相对路径,适配不同用户环境

**v0.4.0** (2026-03-15) — 架构重构版本  
- 优化意图识别,语义理解为主,关键词辅助匹配
- 新增三级降级策略  
- 修复 3 个核心 Bug,测试通过率 88.9%

---

*"我有一个穿搭闺蜜,她叫 StyleBuddy"*  
*— 让穿搭更简单,让每天更自信 💃*