Skip to content
Development
Skill

/cultivating-personas

Distills AI agent voice patterns from conversation into a Persona Voice Card v1.0 (self/user/language/register/emoji_policy/flourish only — no judgment-shaped fields), validates schema + content safety + differentiation, and routes contributions through the existing submission

From plugin
code-abyss
23830 skills
Install
$ npx -y skills add telagod/code-abyss --skill cultivating-personas --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/cultivating-personas

Context preview

The summary Claude sees to decide when to auto-load this skill.

Distills AI agent voice patterns from conversation into a Persona Voice Card v1.0 (self/user/language/register/emoji_policy/flourish only — no judgment-shaped fields), validates schema + content safety + differentiation, and routes contributions through the existing submission

SKILL.md

cultivating-personas.SKILL.md
name: cultivating-personas
description: Distills AI agent voice patterns from conversation into a Persona Voice Card v1.0 (self/user/language/register/emoji_policy/flourish only — no judgment-shaped fields), validates schema + content safety + differentiation, and routes contributions through the existing submission portal. Use when the user wants to crystallize a recurring voice/tone, fork an existing persona, or share one with the community.
compatibility: node>=18
user-invocable: true
allowed-tools: Bash, Read, Write, Edit
argument-hint: <distill|create|validate|publish> [persona-slug]
aliases: forge-persona, cultivate-persona

<!-- safety-scan: ignore TOOLS_PRIVILEGED 特权理由:Bash 用于 schema 校验脚本;Write/Edit 用于在用户确认后落盘 <slug>.json -->

炼制人格 · cultivating-personas

> 人格是声音,不是判断。persona-voice-card 只有 self / user / language / register / > emoji_policy / flourish 几个受限字段——没有自由文本字段可以藏判断内容,沉淀和校验 > 因此都比 v1 简单得多:机器能判的就是全部,不再需要"identity 三段是否齐全"这类主观阅读。

三种触发模式

| 模式 | 触发 | 入口 | |------|------|------| | `distill` | 会话中识别"魔尊形成稳定声音模式" | 主动提议→显式确认 | | `create` | 用户显式想造新人格 | 引导式问答→生成 card | | `validate` | 已有 card,校验 schema + 内容安全 + 差异度 | 一键扫描 |

何时使用

| 场景 | 使用 | 理由 | |------|------|------| | 自称 / 称呼用户已稳定 ≥ 3 会话 | ✅ distill | 声音模式可结晶 | | 用户想 fork 内置人格做变体 | ✅ create | 直接生成骨架 | | 已有 card 但质量存疑 | ✅ validate | 客观闸门 |

何时不使用

  • ❌ 单次会话的临时口吻——口吻不是人格
  • ❌ 人格内容含真实人名 / 政治立场——风险高,建议改写
  • ❌ self+user 与现有 6 个内置人格完全相同——用 fork 或 PR 改进而非新建
  • ❌ 仅是改 emoji_policy 或 language——改现有 card,不另起
  • ❌ 想让人格携带判断/授权/优先级内容——那是 `skills/_kernel/` 或对应领域 skill 的事,不是人格的事(见下方"安全脊柱")

安全脊柱

人格虽是文本,仍有红线:

1. **schema 合规(结构性,机器判)**:通过 [persona-voice-card.schema.json](../../docs/specs/persona-voice-card.schema.json) 验证——`additionalProperties:false`,没有自由文本字段能装下判断内容,这是设计上的,不是靠自觉 2. **无政治 / 宗教 / 民族敏感内容**——会因平台审查阻断分发 3. **无真实人名 / 商标盗用**——法律红线 4. **self / user 必须自然**——非自然称谓(单个非中文符号)warn 5. **不含判断性字段**——schema 里根本没有 authorization / scenarios / priority 这类字段,validate 会直接拒绝任何试图塞入的未知字段

详细审查清单见 [references/persona-safety.md](references/persona-safety.md)。

工作流速查

| 你想做的 | 走哪卷 | |---------|--------| | 从会话沉淀人格 | [references/distillation.md](references/distillation.md) | | 校验 voice 差异度 | [references/voice-consistency.md](references/voice-consistency.md) | | 通过 submit portal 提交 | [references/publishing.md](references/publishing.md) |

使用

# 校验现有 card
node scripts/persona_forge.js validate config/personas/<slug>.json

# 从会话提炼(交互式,不直接落盘,先输出预览)
node scripts/persona_forge.js distill --voice-hint "<观察笔记>"

# 生成提交 payload(给 submit.html 用)
node scripts/persona_forge.js publish config/personas/<slug>.json

收口

  • 校验通过 → 交付 `<slug>.json` 给魔尊 review
  • 提交社区 → **复用** [submission portal](https://telagod.github.io/code-abyss/submit.html#zh)
  • 本 skill **不重造**提交流程,只生成 payload + 引导跳转

参见姊妹 skill [cultivating-skills](../cultivating-skills/SKILL.md)——专司工程方法沉淀。

Read more
Ships withcode-abyss

Give your AI coding agent a personality. Composable persona + style + skills for Claude Code, Codex, Gemini CLI & OpenClaw. Ships Tech Persona Card v1.0 spec.

Get the whole plugin

Other skills on code-abyss.