agentleFS
kachofugetsu09kachofugetsu09/akashic-agent/skills/skill-creater/SKILL.md

skill-creater

Skill520 starsChanged 3 months agoMITView on GitHub

创建或改写 akashic-bot 技能(SKILL.md)。当用户要求新建技能、适配现有技能到当前格式、或修改技能内容时使用。

  • Reads credentials

The file

Files
SKILL.md

Skill 创建指南

技能目录结构

$AKASHIC_WORKSPACE/skills/{skill-name}/   ← 工作区技能(优先级最高,无需写插件直接热生效)
  SKILL.md                                ← 必须
  scripts/                                ← 可选,放辅助脚本
  references/                             ← 可选,放参考文档

~/.akashic/skills/{skill-name}/           ← 用户级技能
<Core 根>/skills/{skill-name}/            ← 内置技能
随插件发布的技能                           ← 见 develop-akashic-plugin

用 write_file 或 edit_file 直接创建/修改文件,无需任何脚手架脚本。若要从 GitHub 仓库或本地目录安装现成技能,请加载 skill-installer。

SKILL.md 格式

---
name: skill-name
description: 一句话说明技能功能 + 触发场景。触发词:词1, 词2, 词3
metadata: {"akashic": {"always": false, "requires": {"bins": ["curl"], "env": ["API_KEY"]}}}
---

# 技能标题

正文指令...

frontmatter 字段说明

  • name:技能目录名(小写,连字符)
  • description:最重要的字段。agent 靠它判断何时触发此技能,务必包含功能描述和触发词
  • metadata:运行配置
    • always:true = 每轮对话都注入完整正文;false = 按需加载(默认)
    • requires.bins:所需 CLI 工具,缺失时技能标记为不可用
    • requires.env:所需环境变量,缺失时技能标记为不可用

写作原则

简洁优先:SKILL.md 会被注入到 context,每个 token 都有成本。

  • 只写 agent 不已知的信息
  • 用示例命令代替大段文字解释
  • 把冗长的参考内容放到 references/ 子文件,在 SKILL.md 里注明"详见 references/xxx.md"
  • 正文控制在 100 行以内

触发词要全:中英文都写,覆盖用户可能说的各种表达方式。

创建流程

  1. 确认技能名称和保存位置(默认工作区;要随插件发布时改用 develop-akashic-plugin)
  2. 用 write_file 写入 SKILL.md
  3. 如有辅助脚本,写入 scripts/ 目录
  4. 验证:requires.bins 里的命令是否正确

改写已有技能

改写时重点检查:

  • frontmatter 是否有 metadata 字段(旧技能可能缺失)
  • requires.bins 是否列全了依赖
  • 正文是否有指向不存在脚本的引用(如 init_skill.py)
  • 安装命令是否适配当前系统(Linux 用 uv/pacman/apt,而非 brew)

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.