Skip to main content

Skill Display Name Generation

@ aquarius-wing1632026.7.10skill-name-generation

Generate paired English + Chinese display names for an existing skill — 1 recommended + 3 alternates in a single pass, candidates only, no files touched.

FeaturedAI Engineering

Basic info

Name
Skill Display Name Generation
Description
给一个已有 skill 的 slug + description 生成人类可读的显示名, 一次配对输出 `display_name` (英文 Title Case 短语) 与 `display_name_zh` (中文 4-8 字), 用于 frontmatter / 展示层。触发于: "给这个 skill 起个显示名 / 起个中文名 / 起个人类可读名 / 生成 display_name / display name / skill 叫什么 / skill 起名 / skill 命名 / human-readable name / 帮我给 skill X 起个中英文名"。也覆盖批量回填存量 skill 显示名字段的场景。Do NOT trigger when: 起产品/功能/模块名 (走 product-title-generation) / 给代码变量·函数·类命名 (走 code naming) / 起 skill 的 slug 容器名 (走 skill-domain-framing) / 写 slogan·营销文案·SEO 标题·商标合规。也不改任何 SKILL.md 或 openai.yaml 文件——写不写字段由用户自行决定。

skill-name-generation

Generate paired English + Chinese display names for an existing skill — 1 recommended + 3 alternates in a single pass, candidates only, no files touched.

When to use it

Name one skill:

You just finished the slug and description for a skill and need to fill in display_name / display_name_zh in its frontmatter — you want paired candidates in one shot.

Name several skills:

A batch of older skills in the repo has slugs but no display names. You want the same conventions applied across the board, English and Chinese paired without drift.

Regenerate display names:

An existing display name reads like an explanation, not a name; or the word you'd reach for when trying to find this skill later doesn't appear in it. You want to see whether a tighter candidate exists.

Path-only input:

You don't want to copy-paste the slug and description — hand over a SKILL.md path and let the skill pull them from the frontmatter itself.

Out of scope:

Naming a product / feature / module → product-title-generation; naming a skill's slug (lowercase, hyphenated) → skill-domain-framing; deciding whether a piece of content is worth crystallizing into a skill → skill-content-fit; code variables / functions / classes, ad slogans / marketing copy / SEO titles / trademark clearance → don't trigger.

What it produces

Candidate text only — never edits any file. Whether and where to write is your call. This is the most counterintuitive part.

  • Paired in one pass: 1 recommended + 3 alternates. Each set = display_name (English, Title Case phrase) + display_name_zh (Chinese, 4–8 characters) + one line of rationale (must call out which phrase in the description the core word was pulled from).
  • Never does: touch the target skill's SKILL.md / agents/openai.yaml / index.json or any other repo file.

Prerequisites & boundaries

Prerequisites:

Needs two pieces of information — the skill's slug and its description. Alternatively, hand over a SKILL.md path and the skill will pull name and description from the frontmatter itself (body ignored). If both are missing, ask once; if still missing, stop.

Adjacent skills:

Task Route to
Name a skill's slug (lowercase, hyphenated) skill-domain-framing
Name a short Chinese title for a product / feature / module product-title-generation
Orchestrate the full skill creation flow skill-create-workflow

Out of scope:

  • Code variable / function / class naming
  • Ad slogans / marketing copy / SEO titles / trademark clearance
  • Bootstrapping a skill from scratch (that's skill-create-workflow)

Subtle edges:

  • Target skill already has a display_name → ask once "regenerate or keep", default to keep if the user doesn't confirm
  • Slug suffixes (-workflow / -set / -review) are only auxiliary signals; the primary job is always judged by the first capability verb in the description

Let's Build Together

Follow us and join the community for updates

WeChat community

Scan to join WeChat group

WeChat QR Code