Back to Details

skill-creator

Compare original and translation side by side

🇺🇸

Original

English
🇨🇳

Translation

Chinese

Skill Creator

Skill 创建器

Build skills that behave like “mini senior engineers”: trigger correctly, run deterministically, and leave resumable artifacts on disk.
构建表现如同「迷你资深工程师」的Skill：触发准确、运行可预测，并在磁盘上留下可恢复的产物。

What good looks like (Ship Faster standard)

优秀Skill的标准（Ship Faster规范）

  • SKILL.md
    : short and triggerable (routing + safety + I/O + output contract). Avoid long tutorials here.
  • references/
    : deep docs loaded only when needed (progressive disclosure). Link directly from 
    SKILL.md
    .
  • scripts/
    : deterministic helpers for repetitive/fragile work (prefer this over re-writing code in chat).
  • assets/
    : templates or files to copy (not meant to be loaded into context).
Deep guidance:
  • Progressive disclosure + authoring principles: 
    references/skill-design-principles.md
  • Workflow structure patterns: 
    workflows.md
  • Output templates/patterns: 
    output-patterns.md
  • SKILL.md
    ：简洁且可触发（包含路由+安全+I/O+输出契约）。此处避免冗长教程。
  • references/
    ：仅在需要时加载的深度文档（渐进式披露）。直接从
    SKILL.md
    中链接至此处。
  • scripts/
    ：用于重复/易出错工作的可预测辅助脚本（优先使用此方式，而非在对话中重写代码）。
  • assets/
    ：用于复制的模板或文件（不加载至上下文环境中）。
深度指导：
  • 渐进式披露与创作原则：
    references/skill-design-principles.md
  • 工作流结构模式：
    workflows.md
  • 输出模板/模式：
    output-patterns.md

Creation flow (recommended)

推荐创建流程

  1. Collect trigger examples (real user phrases that should activate the skill).
  2. Define the I/O contract (paths only) + required artifacts:
    • What files does the skill read?
    • What files must it write (plans, evidence, summaries)?
  3. Implement reusable resources first:
    • add 
      scripts/
      for deterministic steps
    • add 
      references/
      for heavy documentation
    • add 
      assets/
      for templates / boilerplate
  4. Write 
    SKILL.md
    last    :
    • frontmatter 
      description
      should include trigger phrases
    • body should be short and link out to the resources above
  5. Validate + package (optional) using the scripts below.
  1. 收集触发示例（能够激活该Skill的真实用户语句）。
  2. 定义I/O契约（仅包含路径）及所需产物：
    • Skill需要读取哪些文件？
    • 它必须写入哪些文件（计划、证据、摘要）？
  3. 优先实现可复用资源：
    • 添加
      scripts/
      目录存放可预测步骤脚本
    • 添加
      references/
      目录存放深度文档
    • 添加
      assets/
      目录存放模板/样板代码
  4. 最后编写
    SKILL.md
    ：
    • 前置元数据
      description
      需包含触发语句
    • 正文应简洁，并链接至上述资源
  5. 验证与打包（可选）：使用下方脚本完成。

Scripts

脚本说明

All scripts live in 
scripts/
in this skill folder.
所有脚本均位于此Skill目录的
scripts/
文件夹中。

Initialize a new skill folder

初始化新Skill目录

bash
python scripts/init_skill.py <skill-name> --path <output-directory>
bash
python scripts/init_skill.py <skill-name> --path <output-directory>

Validate a skill (frontmatter + basic conventions)

验证Skill（元数据+基础规范）

bash
python scripts/quick_validate.py /path/to/<skill-name>
bash
python scripts/quick_validate.py /path/to/<skill-name>

Sync catalogs/assets (manifest, skills-map, docs catalog)

同步目录/资源（清单、Skill映射表、文档目录）

bash
python scripts/sync_catalog.py
bash
python scripts/sync_catalog.py

Lint skills (compat + quality gates)

检查Skill兼容性与质量

bash
python scripts/skill_lint.py --check-generated
bash
python scripts/skill_lint.py --check-generated

Package a skill into a distributable 
.skill
zip (optional)

将Skill打包为可分发的
.skill
压缩包（可选）

bash
python scripts/package_skill.py /path/to/<skill-name> [output-directory]
bash
python scripts/package_skill.py /path/to/<skill-name> [output-directory]

Notes for Ship Faster contributors

Ship Faster贡献者注意事项

  • Prefer prefix naming: 
    workflow-
    , 
    tool-
    , 
    review-
    , 
    mcp-
    , 
    skill-
    , 
    publish-
    .
  • Keep side effects gated: write a plan first, ask for explicit confirmation, then execute, then verify.
  • If you add/remove a skill, update the catalog: 
    skills/manifest.json
    .
  • 优先使用前缀命名：
    workflow-
    , 
    tool-
    , 
    review-
    , 
    mcp-
    , 
    skill-
    , 
    publish-
    。
  • 副作用操作需受控：先编写计划，请求明确确认后再执行，最后进行验证。
  • 若添加/移除Skill，请更新目录：
    skills/manifest.json
    。

Claude Code: Slash Commands merged into Skills

Claude Code：斜杠命令与Skill合并

Claude Code now treats custom slash commands and skills as the same concept:
  • .claude/commands/<name>.md
    and 
    .claude/skills/<name>/SKILL.md
    both create 
    /<name>
  • Existing 
    .claude/commands/
    files keep working (no migration required)
  • Prefer creating skills going forward so you can bundle 
    references/
    , 
    scripts/
    , and use subagents
Useful frontmatter knobs (Claude Code):
  • disable-model-invocation: true
    - only user can invoke (good for side-effect workflows)
  • user-invocable: false
    - hide from users; Claude-only background knowledge
  • context: fork
    + 
    agent: Explore|Plan|...
    - run skill in a subagent (good for research)
  • allowed-tools
    - restrict tool surface when this skill is active
Claude Code 现在将自定义斜杠命令与Skill视为同一概念：
  • .claude/commands/<name>.md
    和 
    .claude/skills/<name>/SKILL.md
    均可创建
    /<name>
    命令
  • 现有
    .claude/commands/
    文件可继续使用（无需迁移）
  • 未来优先创建Skill，以便捆绑
    references/
    、
    scripts/
    并使用子Agent
Claude Code 实用元数据配置项：
  • disable-model-invocation: true
    - 仅用户可触发（适用于有副作用的工作流）
  • user-invocable: false
    - 对用户隐藏；仅作为Claude的背景知识
  • context: fork
    + 
    agent: Explore|Plan|...
    - 在子Agent中运行Skill（适用于研究场景）
  • allowed-tools
    - 限制该Skill激活时的工具使用范围

Archived full guide

归档完整版指南

The previous (long-form) version of this skill is preserved at:
  • references/legacy-skill-creator.md
本Skill的旧版（长篇）指南已存档于：
  • references/legacy-skill-creator.md