面部顾问

安装指南

1. ClawHub CLI 从源直接安装技能的最快方式。
   npx clawhub@latest install youcam-facial-consultant
2. 手动安装 将技能文件夹复制到以下位置之一
   全局模式
   ~/.openclaw/skills/
   工作区
   /skills/
   优先级：工作区 > 本地 > 内置
3. 提示词安装 将此提示词复制到 OpenClaw 即可自动安装。
   请帮我使用 Clawhub 安装 youcam-facial-consultant。如果尚未安装 Clawhub，请先安装（npm i -g clawhub）。
   更倾向于下载？
   获取包含原生技能文件的ZIP压缩包。
   下载技能 (.zip)

什么是

面部顾问是一种由 YouCam（也称为 Perfect Corp AI）提供支持的 Openclaw Skills 图像分析工作流。它通过 scripts/youcam_core.py 运行用户选择的肤色分析和面部属性分析 API，生成一份易读的报告，介绍面部颜色、五官形状、脸型和面部比例。
此技能旨在提供化妆品和外观参考，而非医疗诊断、皮肤状况评分、化妆建议或试戴发型。它会先让用户选择分析类型，告知积分费用，验证 API 凭据，使用最新可用的 API 规范，并清楚标注近似结果或缺失结果。

应用场景

从清晰的自拍中识别返回的肤色、眼睛、眉毛、嘴唇和头发颜色。
分析眼型、眼睛大小、眼距、眉形、颧骨、脸型、唇形、鼻部尺寸和面部比例等面部属性。
查看横向三庭、纵向五眼、脸部长宽比，以及上下唇比例。
仅在用户明确同时选择两项时，运行颜色分析和面部属性分析。
提供仅供参考的非医疗面部信息，避免给出缺乏依据的底色或整形手术结论。
将皮肤状况评分请求转交给独立的 skin-analysis-expert 技能。

工作原理

判断用户需要肤色分析、面部属性和比例分析，还是两者都需要；说明每项所选分析的积分费用。
从相关 Perfect Corp API 文档中获取当前端点和参数详情，与 api-fallback.yaml 进行比较；如果有更新的 API 版本，则传入 --version。
确认用户提供的是清晰、正面且仅包含一人的自拍。
验证 YOUCAM_API_KEY，且不消耗积分。
仅通过 scripts/youcam_core.py 运行用户批准的分析。
如果 API 返回正在运行的任务，则保留返回的 task_id 并轮询其状态，而不是启动重复且会产生费用的任务。
阅读 references/guide.md 以解读颜色和特征值，并使用指南中的目标值计算黄金比例区间。
省略不可用的数值，将比例描述为近似值，并避免推断 API 未返回的底色。
将结果整理成易读的报告，并始终附上必需的 AI 分析免责声明。

配置指南

从 https://yce.perfectcorp.com/api-console 获取 Perfect Corp YouCam API 密钥。
通过环境变量提供密钥，或将其放入技能根目录（与 SKILL.md 同级）的 credentials.json 文件中。
确认可以访问网络 yce-api-01.makeupar.com，并拥有执行 Python 的权限。
发起付费请求前验证凭据：
python scripts/youcam_core.py validate-key
检查每项可选分析当前所需的积分，并在运行前告知用户：
python scripts/youcam_core.py cost --feature skin-tone-analysis
python scripts/youcam_core.py cost --feature face-attr-analysis
仅对清晰的自拍运行用户选择的功能：
python scripts/youcam_core.py run --feature skin-tone-analysis --src_file --param face_angle_strictness_level=high python scripts/youcam_core.py run --feature face-attr-analysis --src_file --param face_angle_strictness_level=high --param features='["eyeShape","eyeSize","eyeDistance","eyebrowShape","cheekbones","faceShape","lipShape","noseWidth","noseLength","eyeColor","lipColor","eyebrowColor","hairColor","horizontalThird","verticalFifth","faceAspectRatio","upperLipToLowerLip"]' 如果运行结果中的 task_status 为 running，请勿重新运行。应在最长 24 小时内轮询现有任务：
python scripts/youcam_core.py status --feature --task_id
如果未配置 API 密钥，请引导用户前往 Perfect Corp API 控制台，并提供环境变量或 credentials.json 配置方式。

数据架构

配置和参考文件
项目 用途
SKILL.md 定义触发条件、工作流、API 要求、安全规则和输出行为。
scripts/youcam_core.py 验证凭据、检查费用、启动分析并轮询异步任务。
credentials.example.json 说明本地凭据文件的预期格式。
credentials.json 可选的 YOUCAM_API_KEY 本地存储文件；请放在技能根目录。
references/guide.md 提供解读指南、黄金比例目标值和分档规则。
api-fallback.yaml 最新文档不可用时使用的备用 API 规范。
分析分类
skin-tone-analysis：返回肤色、眼睛、眉毛、嘴唇和头发颜色数据。
face-attr-analysis：返回面部特征形状、脸型、所选面部尺寸和比例测量值。
task_id：标识异步 API 任务，有效期为 24 小时。
task_status：跟踪处理状态，包括 running、success 或 error。
features：指定向面部分析器请求哪些面部属性的 JSON 数组。
黄金比例解读：根据返回的测量值，使用 references/guide.md 中的分档和目标值在本地计算。
缺失值会被跳过，不会编造；由于每次运行的结果可能不同，比例会以近似值报告。

高级特性

支持两项相互独立的 Perfect Corp AI 分析，由用户选择并关注积分消耗。
每次运行都会获取最新的端点和参数文档，而非仅依赖静态 API 定义。
通过轮询
task_id
安全处理异步任务，防止意外重复扣费。
支持配置面部角度严格度，并使用明确的特征选择数组进行详细面部分析。
references/guide.md
和黄金比例分档，将原始 API 结果与本地解读分开。
支持使用环境变量凭据或技能本地的
credentials.json
配置。
坚持审慎报告：不推断缺乏依据的底色、不编造缺失值，并强制附上非医疗免责声明。
适用于需要进行结构化自拍分析，同时明确用户同意和付费 API 执行的 Openclaw Skills 工作流。

name: facial-consultant description: >- Analyze skin tone and facial attributes from a single selfie using YouCam
(Perfect Corp) AI. Returns skin/eye/eyebrow/lip/hair colors plus facial feature
shapes and golden-ratio proportions, as a readable report. Use for "測膚色",
"臉型分析", "facial attributes". Do NOT use for skin-condition scoring (that's
skin-analysis-expert), makeup, or hair try-on.
version: 1.1.1
requirements:
credentials:

• name: YOUCAM_API_KEY source: env | credentials.json (in the skill root, next to SKILL.md)
  network:
• yce-api-01.makeupar.com apis:
• skin-tone-analysis
• face-attr-analysis permissions:
• type: exec commands: [python]

Facial Consultant

An analysis skill that can call two APIs. You (the agent) run the flow here through scripts/youcam_core.py, and you also do the golden-ratio interpretation from the references.

When to trigger

When the user wants their skin tone, facial features / face shape, or facial proportions. For skin-condition scores, hand off to skin-analysis-expert.

Ask first (this skill has multiple analyses)

Before running anything, ask the user what they want and state the credit cost of each:

Skin tone — colors of skin/eyes/brows/lips/hair (skin-tone-analysis)
Facial attributes & ratios — feature shapes, face shape, golden-ratio proportions (face-attr-analysis)
Both Get the per-call cost with python scripts/youcam_core.py cost --feature skin-tone-analysis and --feature face-attr-analysis, tell the user, and only run the ones they choose.
Get the latest API spec (do this every run)

Read the doc section(s) for the chosen analyses and open the latest version for the current endpoint / parameters:

Skin tone: https://docs.perfectcorp.com/reference/ai_skin_tone_analysis.md
Facial attributes / ratio: https://docs.perfectcorp.com/reference/ai_face_analyzer.md If newer than api-fallback.yaml, pass --version.

Prerequisites

A clear, front-facing, single-person selfie.
Run python scripts/youcam_core.py validate-key (0 credits) to confirm the key. If it is not set, tell the user to get one at https://yce.perfectcorp.com/api-console and set the YOUCAM_API_KEY env var, or create credentials.json in the skill root (see credentials.example.json for the format).
Run (only the chosen analyses)

python scripts/youcam_core.py run --feature skin-tone-analysis --src_file

--param face_angle_strictness_level=high

python scripts/youcam_core.py run --feature face-attr-analysis --src_file

--param face_angle_strictness_level=high
--param features='["eyeShape","eyeSize","eyeDistance","eyebrowShape","cheekbones","faceShape","lipShape","noseWidth","noseLength","eyeColor","lipColor","eyebrowColor","hairColor","horizontalThird","verticalFifth","faceAspectRatio","upperLipToLowerLip"]'

If run prints "task_status": "running", the task is still processing and run stopped waiting. Do not run it again, since that starts a new task and charges again. Check back with python scripts/youcam_core.py status --feature --task_id , using the task_id from that output, until task_status is success or error. A task_id stays valid for 24 hours.

Interpret colors and features with references/guide.md, and do the golden-ratio bucketing yourself (buckets and golden targets are in that file).

Output rules

For colors, report colors only; do not infer an undertone that was not returned.
Ratios drift slightly between runs and faceShape may not be returned; label ratios as approximate and skip missing values instead of filling them in.
Always end with this note: "This is an AI-based facial analysis for reference only. It is not a medical or cosmetic-surgery assessment; results can vary with lighting and pose."
PatentMax 中国专利审查意见通知书答复

Picture Storybook Generator