Skill Authoring logo

Skill Authoring

CommunityPopular
uvwt
skill-authoring

创建、设计、修改、重构和验证 AgentDock Skill 时使用;负责 Agent Skills 兼容的 SKILL.md、可移植核心、引用、辅助脚本、测试、安全边界和本地真实验证。

Overview

Publisheruvwt
Repositoryagentdock
Skill nameskill-authoring
Stars
1.2K
Forks
144
Bundled files
3
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by uvwt on GitHub. Read the source before you install it.

Installation

Install the Skill Authoring AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/uvwt/agentdock/tree/main/core-skills/skill-authoring
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/uvwt/agentdock.git /tmp/agentdock
mkdir -p .claude/skills
cp -r /tmp/agentdock/core-skills/skill-authoring .claude/skills/skill-authoring
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Skill Authoring in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Skill Authoring on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Skill Authoring is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Skill Authoring

用于创建或维护 AgentDock Skill。Skill 的本体是模型可读取的说明文档;真实操作仍由命令、文件、浏览器、MCP 等工具完成。

AgentDock 不为 Skill 建立独立版本生命周期。Skill 的当前内容由来源与内容摘要识别;同名 managed Skill 更新时直接原子替换当前内容,不保留可选历史 revision,不提供 activate 或 rollback。

核心原则

  1. 先定义模型何时应该选择该 Skill,再写正文。
  2. Skill 核心应兼容 Agent Skills 的文档模型,不把 AgentDock 私有运行时当作通用契约。
  3. 最小 Skill 只有一个 SKILL.md;只有确有需要时才增加 references/、脚本或测试。
  4. 包内资源使用相对路径,业务环境变量由宿主注入当前进程。
  5. 秘密、缓存、会话、数据库、下载结果和设备私有状态不得进入 Skill 包。
  6. 修改后必须真实 lint、测试,并在需要时安装到 AgentDock 验证当前内容。
  7. 不创建或恢复 Skill 自有版本选择、激活、回滚或历史 revision 机制。

完整规范见 references/skill-package-spec.md。

目录

普通第一方和社区 Skill 默认放在独立 Skill 仓库:

text
skills/<skill-name>/
└── SKILL.md

按需扩展:

text
skills/<skill-name>/
├── SKILL.md
├── references/
├── scripts/
├── run.py
└── tests/

只有必须随 AgentDock runtime 一起交付的核心 Skill 放在:

text
core-skills/<skill-name>/

不要创建空目录。

SKILL.md Frontmatter

AgentDock 使用 Agent Skills 风格的字段:

yaml
---
name: example-skill
description: 清楚说明何时使用、解决什么问题
license: Apache-2.0
compatibility: Requires Python 3.11 or later.
metadata:
  owner: example-team
allowed-tools: exec_command
---

# Example Skill

AgentDock 真正依赖并严格校验的只有:

  • name 必填,长度 1–64,只允许小写 ASCII 字母、数字和 -,不能以 - 开头/结尾,也不能包含连续 --;
  • description 必填,最长 1024 个 Unicode 字符,并能让模型稳定判断何时使用;重要的 Use when、Do not use 和相邻 Skill 边界条件不要依赖正文补充,因为模型会先用 description 做候选路由;
  • Markdown 正文必须非空;
  • 其他 frontmatter(包括 license、compatibility、metadata、allowed-tools、version 以及第三方扩展字段)由作者生态定义,AgentDock 原样保留但不作为安装/运行前提;
  • 不设计 AgentDock 私有的 version、active_version、revision 或 rollback 契约。

AgentDock 路由索引

AgentDock 会先暴露轻量 description 索引,再按需读取完整 SKILL.md。不同来源的索引预算不同:

  • standalone managed 与 Plugin-owned Skill:agentdock_context.skills 返回 trim 后的完整 description;
  • workspace Skill:workspace_context.workspace_skills 返回 trim 后的完整 description,但最多列出 50 项;
  • shared/common Skill:agentdock_context.common_skills 面向数量不可控的 ~/.agents/skills,最多列出 50 项,并把每项 description 限制在 120 bytes。

因此 authoring 时应把 description 视为路由契约,而不是正文摘要的随意前缀。AgentDock 不提供可配置的 description 截断上限;如果未来 Skill 总量显著增长,应通过索引总预算或检索式路由解决,而不是静默裁掉每个已管理 Skill 的 description 后半段。

可移植核心

目标 Skill 默认只假设:

  • 当前进程能读取 Skill 包内容;
  • 包内资源使用相对路径;
  • 环境变量来自当前进程;
  • 宿主提供命令、文件、浏览器或远端工具能力。

有根目录脚本时,通用示例写成:

bash
printf '%s' '{"skill_action":"status"}' | python3 run.py

不要把以下内容作为核心运行前提:

  • ~/.agentdock/skills/<name>;
  • AGENTDOCK_HOME、AGENTDOCK_SKILL_DIR 等 AgentDock 私有路径变量;
  • skill_ref、exec_command 或 skill://;
  • 固定用户绝对路径;
  • 主动读取或 source AgentDock 私有环境文件。

AgentDock 专属说明可以放在独立的“AgentDock 适配/验证”章节;删除该章节后,核心流程仍应成立。

环境变量

需要配置时,在正文中明确声明变量名、类型、必填性和用途:

markdown
## 环境变量

| 变量 | 类型 | 必填 | 说明 |
|---|---|---:|---|
| EXAMPLE_BASE_URL | config | 是 | 服务地址 |
| EXAMPLE_API_KEY | secret | 是 | API Key |

Skill 只声明变量,不保存真实值。辅助脚本只从当前进程环境读取。

AgentDock 本地验证时:

  • skill_manage env_list 查看变量名称与配置状态;
  • skill_manage env_set 写入用户明确提供的值;
  • skill_manage env_unset 删除指定变量;
  • exec_command skill_ref=<host-issued-ref> 运行时只把 managed Skill 环境注入对应子进程。

环境值不写入 AgentDock 主进程或系统全局环境。

需要持久可变状态时,不要写 Skill 包目录。AgentDock 对 standalone managed 与 Plugin-owned Skill 的 exec_command 都会提供独立运行时保留变量 SKILL_DATA_DIR:

  • standalone managed 指向 ~/.agentdock/data/skills/<name>/;
  • Plugin-owned 指向 ~/.agentdock/data/skills/.plugin/<plugin>/<skill>/,并额外获得 Plugin 共享兼容目录 PLUGIN_DATA_DIR=~/.agentdock/data/plugins/<plugin>/;
  • 数据目录只在对应 Skill / Plugin 运行时真正需要时创建;
  • Unix 权限收紧为 0700,Windows 使用当前用户私有 ACL;
  • Windows 原生命令收到 Host 路径;WSL 命令收到已转换的 Linux 路径;
  • skill_manage env_set、宿主 env mapping 和 request.env 都不能覆盖运行时保留变量;
  • Plugin-owned Skill 的用户环境隔离在 ~/.agentdock/env/skill/plugin/<plugin>/<skill>.env;standalone 仍使用 ~/.agentdock/env/skill/<name>.env;
  • shared/workspace 候选不会得到 SKILL_DATA_DIR 或 PLUGIN_DATA_DIR。

SKILL_DATA_DIR / PLUGIN_DATA_DIR 是 AgentDock 可选适配,不是 Agent Skills 通用前提。可移植 Skill 不应把它们列为用户必填配置;需要状态目录的辅助脚本可以在检测到它们时优先使用,并在其他宿主下采用自己明确声明的可移植策略。

引用与辅助脚本

正文引用包内文件时使用相对路径,例如:

text
references/api.md

辅助脚本应:

  • 通过 stdin 接收 JSON 对象;
  • 顶层动作字段使用 skill_action;
  • secret 只从环境读取;
  • 输出结构化 JSON;
  • 错误返回稳定 code 和可读 message;
  • 默认状态检查只读;
  • 破坏性动作要求明确确认;
  • 不回显 secret;
  • 用相对路径或脚本目录定位包内只读资源;
  • 可变缓存、数据库、会话和下载结果不得写回 Skill 包;在 AgentDock managed 运行时优先写入 SKILL_DATA_DIR。

Authoring lint

本 Skill 的 run.py 提供:

  • status:返回 lint_version 与规则数量;
  • lint:检查可移植性和明显宿主绑定。

示例:

json
{
  "skill_action": "lint",
  "source": "/path/to/skills/example-skill"
}

硬错误包括:

  • 硬编码 AgentDock managed Skill 安装目录;
  • 依赖 AgentDock 私有目录变量;
  • 主动读取 AgentDock 环境文件;
  • 固定用户绝对路径。

AgentDock 专属 skill_ref、exec_command、skill:// 出现在目标 SKILL.md 中会作为 warning,要求确认它们只存在于可选宿主适配说明。

安全与质量检查

提交前至少检查:

  • 无真实密码、Token、Cookie、私钥和认证缓存;
  • 无 .env、数据库、截图、下载结果、__pycache__、node_modules 等运行产物;
  • 无符号链接、路径逃逸、固定用户绝对路径;
  • 网络、写入、删除、上传、权限变化和依赖安装均有明确说明;
  • 未恢复 agentdock.yaml、skill_run、skill_env_manage、旧式 operation/entrypoint 清单或统一 Skill Runtime;
  • 不把安装历史、激活版本或回滚机制重新写回 Skill。

AgentDock 本地验证

AgentDock 的 managed Skill 当前布局是:

text
~/.agentdock/skills/<name>/
  SKILL.md
  ...

这个目录是宿主实现细节,不应硬编码进目标 Skill。

建议验证流程:

  1. 检查目录、Frontmatter 和正文;
  2. 检查禁止文件、符号链接和真实 secret;
  3. 对辅助脚本做语法检查和自带测试;
  4. 运行 skill-authoring lint,确认 portable=true;
  5. 使用 skill_manage install 从本地目录、压缩包或 HTTPS 来源安装;
  6. 记录返回的 content_digest;相同内容重复安装应为 no-op;
  7. 通过 agentdock_context 找到该候选,并保存它返回的 skill_ref 与 file;
  8. 用 read_file 读取宿主返回的 file,不要自己按名称重建 URI;
  9. 有辅助脚本时,用同一候选的 skill_ref 调用 exec_command 做只读检查;
  10. 修改内容后再次安装,确认当前内容被原子替换且环境/数据未被清空。

managed、shared、workspace 中同名 Skill 是不同候选。验证时始终使用宿主返回的精确 skill_ref,不得靠裸名称重新解析。

完成标准

只有同时满足以下条件才算完成:

  • description 可稳定触发且职责边界清楚;
  • Frontmatter 与 Agent Skills 文档模型兼容;
  • 可移植核心不依赖 AgentDock 私有目录或工具参数;
  • 包内脚本可从 Skill 根目录按相对路径运行;
  • lint、测试和安全检查通过;
  • managed 安装时有有效 content_digest,重复相同内容为 no-op;
  • agentdock_context 返回正确来源、skill_ref 与 file;
  • 当前内容和代表性只读行为已真实验证;
  • 没有 Skill 自有版本选择、激活、回滚或历史 revision 语义。

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Skill Authoring AI skill do?

创建、设计、修改、重构和验证 AgentDock Skill 时使用;负责 Agent Skills 兼容的 SKILL.md、可移植核心、引用、辅助脚本、测试、安全边界和本地真实验证。

Why use Skill Authoring on TypingMind?

Because you install it once and use it with any model. Skill Authoring is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Skill Authoring in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/uvwt/agentdock/tree/main/core-skills/skill-authoring. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Skill Authoring?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Skill Authoring?

As many as you like. As long as a model supports skills, you can use Skill Authoring with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Skill Authoring AI skill free?

Yes. It is published on GitHub by uvwt under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇