技能包创建者

人工智能 社区

技能解读

当前原文的自动摘译,尚未人工审校。

概述

将成功的任务转化为可重复使用的技能包。提取稳定的工作流程,确定本地技能与包级提示中的内容,生成包结构,并仅在工作流程明确且可重复后才将其打包。

1. 规范化源任务

将已完成的任务简化为干净的执行规范:捕获用户目标、具体可交付成果和最终成功的工作流程(而不是完整的探索性记录)。列出所需的技能、工具、文件、秘密和环境假设。将确定性步骤与启发式步骤分开;消除死角和调试噪音。如果任务仍然太宽泛,请缩小范围,而不是写一个模糊的超级技能。如果关键的成功条件取决于隐藏的人类判断,请将包标记为尽力而为的助理工作流程。

2. 决定包装中应包含什么

本地技能(skills/):可重用的程序知识。保持脚本最少,除非再现性取决于精确的文件生成或重复的 shell 步骤。脚本 (scripts/):重复的 shell 或文件生成逻辑,其中可靠性很重要。参考资料 (references/):详细的架构、API 注释或不应使SKILL.md 臃肿的约定。提示 (skillpack.json):UI 的 1-3 个包级入门输入 — 不是 DAG 或状态机。有关确切的包语义,请参阅references/skillpack-format.md。

3. 创建包规范

在写入文件之前,定义包规格。更喜欢一项本地协调员技能加上少量外部技能。最小清单示例:

4. 创建本地协调器技能

使用 frontmatter 和命令式工作流程说明创建skills/<skill-name>/SKILL.md:将稳定的工作流程编写为正文中的命令式步骤。仅在脆弱或重复操作时添加scripts/;添加references/仅用于详细信息。

流狐档案 运行与权限由流狐静态检查推断,不代表作者声明或实际测试结果。
领域
人工智能
作者声明的智能体
未找到明确声明;不据此推断已兼容或已测试
静态检查
88 / 100 · 启发式扫描,不代表运行安全
运行环境证据
未执行;当前没有隔离运行测试证据
作者 / 版本 / 许可证
@CreminiAI · 未声明许可证
上下文用量评级
低消耗
接入复杂程度
需简单配置
是否需要外部接口密钥
不需要
兼容的系统
macOS · Linux · Windows
运行环境要求
未检出明确要求
检测到的文件与系统行为
  • 只读
  • 允许写入 / 修改
  • 命令执行
网络行为
仅限本地
安装命令数
无(仅作为资料)

需要注意: 未声明工具权限范围;实际权限由运行环境决定。

作者说明文件(原文) SKILL.md

# Skillpack Creator

## Overview

Turn a successful task into a reusable SkillPack. Extract the stable workflow, decide what belongs in a local skill versus pack-level prompts, generate the pack structure, and package it only after the workflow is explicit and repeatable.

## Workflow

### 1. Normalize the source task

Reduce the finished task into a clean execution spec:

- Capture the user goal, concrete deliverable, and the final successful workflow (not the full exploratory transcript).
- List required skills, tools, files, secrets, and environment assumptions.
- Separate deterministic steps from heuristic steps; remove dead ends and debugging noise.
- If the task is still too broad, narrow the scope instead of writing a vague mega-skill. If key success conditions depend on hidden human judgment, mark the pack as a best-effort assistant workflow.

Ask for missing stable facts or infer only the low-risk pieces.

### 2. Decide what the pack should contain

- **Local skill** (`skills/`): reusable procedural knowledge. Keep scripts minimal unless reproducibility depends on exact file generation or repetitive shell steps.
- **Scripts** (`scripts/`): repeated shell or file-generation logic where reliability matters.
- **References** (`references/`): detailed schemas, API notes, or conventions that should not bloat `SKILL.md`.
- **Prompts** (`skillpack.json`): 1–3 pack-level starter inputs for the UI — not a DAG or state machine. See `references/skillpack-format.md` for exact pack semantics.

### 3. Create the pack specification

Before writing files, define the pack spec. Prefer one local orchestrator skill plus a small number of external skills. Example minimal manifest:

```json
{
  "name": "company-research",
  "description": "Research a company and produce a summary report",
  "version": "1.0.0",
  "prompts": ["Research {company} and create a report with financials and competitors"],
  "skills": [
    { "name": "research-orchestrator", "source": "./skills/research-orchestrator", "description": "Orchestrate company research across multiple sources" }
  ]
}
```

### 4. Create the local orchestrator skill

Create `skills/<skill-name>/SKILL.md` with frontmatter and imperative workflow instructions:

```yaml
---
name: research-orchestrator
description: "Orchestrate multi-source company research. Use when the user wants a structured company report covering financials, competitors, and market position."
---
```

- Write the stable workflow as imperative steps in the body.
- Add `scripts/` only for fragile or repeated operations; add `references/` only for detailed information.

### 5. Materialize the pack

Use `scripts/scaffold_skillpack.py` when you have the pack spec:

```bash
# Basic
python3 skills/skillpack-creator/scripts/scaffold_skillpack.py \
  --manifest /tmp/skillpack.json \
  --output /absolute/path/to/output-pack

# With zip
python3 skills/skillpack-creator/scripts/scaffold_skillpack.py \
  --manifest /tmp/skillpack.json \
  --output /absolute/path/to/output-pack \
  --zip
```

The script validates the manifest, writes `skillpack.json`, creates `skills/`, copies `start.sh`/`start.bat` from `templates/`, and optionally runs `npx -y @cremini/skillpack zip`.

### 6. Validate the result

Before handing the pack back, confirm:

- The manifest matches the intended pack scope
- Every declared skill has a valid `name`, `source`, and `description`
- Local skills are present under the target pack's `skills/`
- Starter prompts are concrete enough to reproduce the workflow
- Zip only after the pack runs as a directory

## Output Standard

Produce:

1. A short summary of the stabilized workflow.
2. The target pack structure and skill inventory.
3. The created or updated local skill files.
4. The generated `skillpack.json`.
5. Whether the pack was zipped and where the zip lives.

讨论

基于 GitHub Discussions。登录 GitHub 即可参与讨论、点赞、订阅更新。