Skip to content
cccskills
FreePublished on GitHub

skill-profile-builder

Build an initial account profile from public social media content and guided questions. Organize positioning, style, audience, platforms, preferences, and initial insights into six files, marking where information came from and what still needs clarification.

OriginalShow the original description

首次使用引导:从社媒链接分析、从零生成账号画像 Profile。收集社媒链接+运营意图,分析已发内容与收藏喜好,生成 6 维 Profile。当用户说 创建画像/第一次用/帮我建个人设/分析我的账号建画像/从链接生成画像 时使用。

与 skill-profile-manager 的区别:本 SKILL 专做首次从零分析生成;已有画像的编辑/记忆更新/导出/对比/切换找 profile-manager。

See how to install

Good for

  • Creating an account profile for the first time
  • Reviewing positioning and style from existing posts
  • Organizing content preferences and boundaries
  • Identifying missing information in an account profile

Files included (4)

  • SKILL.md3.9 KB
  • EASEL-META.md814 B
  • references/onboarding-questions.md1.9 KB
  • references/social-link-analysis.md1.9 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Profile 构建器(首次引导)

把"空白/半空的画像"变成"可用的 6 维 Profile"。用户首次使用 Easel 时先跑这个:收集社媒链接 + 运营意图 → 分析已发内容/收藏喜好 → 生成 profiles/<名>/ 下 6 个维度文件。生成后交给 account-diagnosis 做起号诊断。

与 profile-manager 的区别

  • profile-builder(本 SKILL) = 从 0 到 1 生成画像(引导填写 + 社媒分析 + 综合成稿)。
  • profile-manager = 已有画像的 CRUD、切换、字段增删改。

输入

用户按需提供(缺的通过追问补,不编造):

项说明
画像名一个人设 = 一个画像(非一个平台),如"科技数码达人"
社媒链接各平台主页 URL(小红书/抖音/B站/微博/知乎/公众号),用于分析已发内容与收藏喜好
运营意图想做什么方向、为什么做、自己的特点/优势/资源
内容偏好喜欢看什么类型、想产出什么类型、参考/对标的账号
红线明确不做的内容、合规底线

输出

profiles/<画像名>/ 下生成 6 个文件(结构见 profiles/_template/): identity.md / style.md / audience.md / platforms.md / preferences.md / memory.md。 每个字段标注来源:[用户自述] / [链接分析] / [待补充]。

执行步骤

  1. 确定画像名,检查 profiles/<名>/ 是否已存在(存在则问是覆盖还是完善)。不存在则复制 profiles/_template/ 为骨架。

  2. 收集社媒链接。按 onboarding-questions.md 的清单引导用户给出主页链接与基础信息。

  3. 分析社媒链接。按 social-link-analysis.md 用 web_fetch 抓取可获取的公开内容(近期发帖标题/题材/互动、公开收藏/点赞),推断:常发题材、风格语气、受众画像、高互动内容特征。

    • 抓取受限(多数平台反爬)时降级:明确告诉用户哪些没抓到,改由用户口述 + 追问补齐,不臆造数据。
  4. 收集运营意图与偏好。按 onboarding-questions.md 逐维度提问:想运营的方向、原因、自己的特点/资源、内容偏好、对标账号、红线。

  5. 综合生成 6 维。把[链接分析]与[用户自述]融合,按 profiles/_template/ 各文件的字段填写:

    • identity ← 定位/差异化/内容方向
    • style ← 语气/开头结构/视觉/节奏/标志元素(链接分析里的高频风格优先)
    • audience ← 核心人群/兴趣/痛点/互动特征
    • platforms ← 各平台账号名/粉丝量级/内容形式(来自链接)
    • preferences ← 要做/不做/合规底线
    • memory ← 首次留空或只放链接分析得到的初步洞察
  6. 标注缺口 + 追问。任何维度信息不足就在该处写 [待补充] 并列出需要用户回答的具体问题(做成 plan 让用户逐条补),不要用通用套话填满。

  7. 写入并回执。写 profiles/<名>/ 各文件,输出一份"画像已生成"摘要:各维度完成度、哪些来自链接分析、哪些待补充,并建议下一步跑 account-diagnosis。

Profile 感知

本 SKILL 是生成 Profile,不消费。生成质量取决于用户提供的链接可抓取程度与自述完整度——信息越全,后续 account-diagnosis 越准。

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

ai-image-gen

FreeSummary

Generate images from written descriptions, transform existing images with editing instructions, or create several variations. Choose image dimensions and quantity, then save the results using an OpenAI-compatible service or apimart with your own API key.

  • Generating an illustration from a written description
  • Changing an image's background using instructions
  • Creating several variations of an existing image
ZJU-REAL/Easel4KUpdated Oct 23, 2026

ai-music

FreeSummary

Generate background music for short videos, vlogs, and social posts from descriptions of mood, style, and instruments. Download MP3 audio, optionally include sung lyrics, and use shared scripts to trim tracks, adjust loudness, or add music to videos.

  • Generating background music for a vlog intro
  • Creating instrumental music to match a social post
  • Generating a sung track from supplied lyrics
ZJU-REAL/Easel4KUpdated Oct 23, 2026

ai-video-gen

FreeSummary

Create new videos from written scene descriptions or still images using external AI services. Set the aspect ratio and duration, animate a character image, and request generated audio when the selected model supports it.

  • Generate a short video from a scene description
  • Animate a still image
  • Create character motion from a starting image
ZJU-REAL/Easel4KUpdated Oct 23, 2026

asset-manager

FreeSummary

Organize generated images, videos, and text in outputs/ by date, platform, and type. Add tags, search filenames and content, produce asset lists and reports, and preview archive moves before applying them to make past work easier to reuse.

  • Organizing generated assets by date, platform, and type
  • Finding past content through keywords and tags
  • Listing assets for a specific account or platform
ZJU-REAL/Easel4KUpdated Oct 23, 2026

audio-denoise

FreeSummary

Reduce background noise in audio recordings or video soundtracks with three processing levels. Keep the original file, create a cleaned version, and receive a report of the settings used and file details before and after processing.

  • Reducing fan or air conditioner noise in recordings
  • Cleaning speech in podcasts and interviews
  • Reducing noise in a video's audio track
ZJU-REAL/Easel4KUpdated Oct 23, 2026

audio-editing

FreeSummary

Edit audio files by trimming sections, converting formats, matching loudness, or extracting sound from videos. You can also join recordings, add fades, and change playback speed while keeping the original pitch.

  • Trim a recording to the section you need
  • Convert audio formats and match loudness
  • Extract audio from a video
ZJU-REAL/Easel4KUpdated Oct 23, 2026

See all skills by ZJU-REAL

Report a problem with this skill