steamer

Copy to AI

• The only entry point of this Skill: current file SKILL.md。
• Pair with Immortality.skill main engine for more complete functionality (collection CLI, character templates, etc.).

Steamer · Steamer

Language

Based on the user'sfirst messagelanguage, use the same language throughout.

When to activate

• The user wants to "distill anyone," "steam a digital avatar," or "use that Steamer setup."
• The user enters from the README or this file and wants to complete the distillation by Phase.

Core concept

Steamer = the narrative entry point of the digital immortality engine. Technical actions:Select character → collect → extract by dimension (procedure / interaction / memory / personality) → grade evidence → package.

Operation sequence

Phase 0: Align expectations

1. Confirm the subject:anyone; if it involves public figures / big names / top influencers, you must state that only public, traceable materialsare used (speeches, papers, interviews, official social posts, etc.).
2. Explain the value in one sentence:use public methodology as an advisor, do not impersonate the person, do not fabricate statements.

Phase 1: Select character template

Confirm the role type of the distillation subject:

你想蒸馏谁？

  [1] 🪞 自己（全维度数字分身）
  [2] 🏢 同事（工作方式与沟通风格）
  [3] 🎓 导师/Mentor（教学方式与指导智慧）
  [4] 🏠 亲人（家族记忆与生活智慧）
  [5] 💕 伴侣/前任（关系记忆与互动模式）
  [6] 🤝 朋友（友谊互动与共同经历）
  [7] 🌐 公众人物（公开方法论）

Ethical emphasis for different roles:

• Colleague/Mentor: limited to internal team alignment and training
• Relative (deceased): confirm whether other family members should know
• Partner/Ex: confirm the purpose is positive remembrance; strictly de-identify
• Public figure: limited to public materials; must have traceable public sources
• Self: note de-identification of others' statements in chats

Phase 2: Collect materials

Provide the following collection methods:

材料怎么提供？

  [A] 自动采集（推荐）
      飞书 / 钉钉 / Slack / Discord / Telegram / Email
      → 扫描频道 → 拉取消息

  [B] 本地数据库
      微信（需第三方导出或本地 SQLite）
      iMessage（macOS，需 Full Disk Access）

  [C] 归档文件
      WhatsApp 导出 / Twitter/X 归档 / Google Takeout
      Facebook 数据下载 / 微博导出

  [D] 上传/粘贴文件
      PDF / JSON / CSV / Markdown / 纯文本

  [E] 直接粘贴文字

可混用多种方式。

Pair with Immortality.skill main engine to use the automatic collection CLI.

Phase 3: Extract by dimension

Determine the required dimensions based on the role type:

Dimension	Applicable roles
Procedural (how to do things)	Colleague, mentor, self, public figure
Interactional (how to speak)	All
Memory (what has been experienced)	Self, relative, partner, friend, mentor, public figure
Personality (what kind of person)	All (colleague only work-related)

Label each output with an evidence level:verbatim(original words) / artifact(document) / impression(impression).

Phase 4: Merge and handle conflicts

Write contradictions into conflicts.md. People are inherently contradictory—do not force unification.

Phase 5: Output the Skill package

Generate the directory structure:

<slug>/
├── SKILL.md          ← AI 读这个就知道「TA 是谁」
├── interaction.md    ← 怎么说话
├── procedure.md      ← 怎么做事（如适用）
├── personality.md    ← 是什么人
├── memory.md         ← 经历过什么（如适用）
├── conflicts.md      ← 矛盾之处
└── manifest.json     ← 元数据

Phase 6: Inform the user

• Generation path and loading method
• Evidence coverage statistics
• Limitation notes for each dimension

What not to do

• Do not fabricate non-public private information; do not encourage unauthorized impersonation of real people.
• Do not treat general common knowledge as the personal characteristics of the person being distilled.
• Do not simulate private conversations of public figures.
• Do not use for stalking, harassment, or deception.

Self-check

• For public figure scenarios, are public sources restricted?
• In each dimension file, verbatim + artifact Does the proportion meet the standard?
• conflicts.md Does it reflect genuine contradictions?
• Does the ethical statement match the role?