Agent skill

Tool Creator

把用户的可复用需求封装成标准化的 Agent-CLI 工具，并用 myagents tool add 注册进 MyAgents 工具注册表——注册后所有未来会话（builtin / DSH / Claude Code / Codex 全 runtime）的 AI 都会在 system prompt 里自动发现它。触发场景：(1) 用户说「把 XX…

AGPL-3.0Auto-check passedAI & LLM Engineering

Install Tool Creator

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .claude/skills && cp -r skills-src/bundled-skills/tool-creator .claude/skills/tool-creator && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Paste into Claude Code

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .claude/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

$skill-installer

$skill-installer install https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator

Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .agents/skills && cp -r skills-src/bundled-skills/tool-creator .agents/skills/tool-creator && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .agents/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/bundled-skills/tool-creator .cursor/skills/tool-creator && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .cursor/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

gemini skills install

$ gemini skills install https://github.com/hAcKlyc/MyAgents.git --path bundled-skills/tool-creator

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/bundled-skills/tool-creator .gemini/skills/tool-creator && rm -rf skills-src

Use ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.

Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/

Paste into Gemini CLI

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .gemini/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator

Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .github/skills && cp -r skills-src/bundled-skills/tool-creator .github/skills/tool-creator && rm -rf skills-src

Use ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.

GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/

Paste into GitHub Copilot

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .github/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

skills CLI

$ npx skills add hAcKlyc/MyAgents --skill tool-creator -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install hAcKlyc/MyAgents tool-creator --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/hAcKlyc/MyAgents.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/bundled-skills/tool-creator .opencode/skills/tool-creator && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "tool-creator" agent skill from https://github.com/hAcKlyc/MyAgents/tree/main/bundled-skills/tool-creator into .opencode/skills/tool-creator/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "tool-creator", then confirm the skill loads.

OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.

Facts

Skill name: tool-creator
GitHub stars: 915
Token cost: ~1.5k tokens
SKILL.md length: 372 words
Files: 3 (incl. references)
Skills in repo: 15
Repo updated: Oct 7, 2026
First seen: Oct 7, 2026
Licence: AGPL-3.0

At a glance

Works in 8 steps: 绝对禁止交互式输入。任何 stdin prompt（确认、选择、密码）都会让… → stdout 只放结果，stderr 放诊断，退出码语义化。 Agent… → 必须有 --json 模式，错误也要结构化。… → …
Tasks that involve Prompt engineering
SKILL.md covers 第 0 步：判断要不要铸工具, 第 1 步：选原型, 第 2 步：写工具——形态契约 and 第 3 步：写三面文档, plus 3 more sections
Calls node; needs API_KEY

What it does

Tool Creator is an agent skill from hAcKlyc/MyAgents. 把用户的可复用需求封装成标准化的 Agent-CLI 工具，并用 myagents tool add 注册进 MyAgents 工具注册表——注册后所有未来会话（builtin / DSH / Claude Code / Codex 全 runtime）的 AI 都会在 system prompt 里自动发现它。触发场景：(1) 用户说「把 XX 封装成工具」「做成一个工具」「注册个工具」「写个 CLI」「以后能直接用」；(2) 用户描述一个会反复出现的自动化需求——文档/文件批量处理、调用某个云 API、用某个多模态大模型做图像/视频理解等，即使没说"工具"两个字；(3) 你发现自己第二次为同类需求写几乎一样的脚本——这时要主动提议把它升格为注册工具，不要等用户开口。反向边界：一次性任务就地解决、不铸工具；接入现成的 MCP server 用 myagents mcp，不归这里。

Its SKILL.md is about 1.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 3 other files, including reference files (for example `references/example-api-tool.md` and `references/example-local-tool.md`).

It sits in AI & LLM Engineering, covering Prompt engineering and MCP servers. It works with Model Context Protocol. The repository describes itself as: MyAgents - 优雅、易用的 Agent 桌面端，一站式 Agent 工作台与任务中心. The licence is AGPL-3.0.

When your agent uses it

Tasks that involve Prompt engineering
Tasks that involve MCP servers

Example prompts

“/tool-creator”

Requirements

A credential in API_KEY

Workflow steps

8 steps, taken from the first numbered list in SKILL.md.

绝对禁止交互式输入。任何 stdin prompt（确认、选择、密码）都会让 Agent 的 shell 调用永久挂死——Agent 没有键盘。危险操作（删除、覆盖、花钱）用确认协议替代：缺 --yes 时打印将要做的变更 + 完整的带 --yes 重跑命令，以退出码 4…
stdout 只放结果，stderr 放诊断，退出码语义化。 Agent 判断成败只靠这三样。退出码约定：0 成功 / 1 一般错误 / 2 用法或参数错误 / 3 环境缺失（缺 env key、缺依赖）/ 4 需要确认。进度提示、调试信息一律 stderr；spinner…
必须有 --json 模式，错误也要结构化。默认输出给人读的简洁文本；--json 输出机器可解析结果。出错时 stderr 给一行 JSON：{"error": "...", "code": "...", "remediation"…
密钥走 env，绝不走 argv。 argv 会泄进进程列表和日志。从 process.env[KEY] 读，缺失时按退出码 3 + remediation 处理。key 在 tool.json::envKeys 声明，由 myagents tool env set…
有界运行时间。不准变 daemon。所有网络请求带 AbortSignal.timeout(...)；轮询循环必须有次数/时间上限。预期超过 30 秒的操作在 readme 里声明耗时量级。
产物文件输出绝对路径到 stdout。生成的文件落到当前工作区（或用户指定路径），把绝对路径作为结果打印——这是下游（预览、IM 发送）能接住产物的前提。
readme 子命令返回标准化使用文档。固定章节（见第 3 步），未来的 AI 第一次用这个工具前会先跑 readme。
依赖自检，失败要可行动。 deps 里声明的二进制在启动时探测（等效 which），缺失则按生死线 3 的错误格式给出安装指引 + 退出码 3——Agent 拿到的是"装 ffmpeg：brew install ffmpeg"，不是一个看不懂的崩溃栈。

What it can do on your machine

Read from SKILL.md and the folder at commit f873ca3. It shows what the files ask for, not the result of running them.

Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
- node
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
- API_KEY
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Tool Creator loads about 1.5k tokens when it runs, and up to ~5.4k if it reads all its reference files. Until then it costs about 103 tokens; SKILL.md has 372 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~103
When it runs · the whole SKILL.md, loaded when a task matches: ~1.5k
With references · SKILL.md plus every file in references/, read only if the agent opens them: ~5.4k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from hAcKlyc/MyAgents at commit f873ca3, republished under its AGPL-3.0 licence (© hAcKlyc). 372 words, ~1,489 tokens.

Download SKILL.mdSave it as .claude/skills/tool-creator/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.

name: tool-creator
description: 把用户的可复用需求封装成标准化的 Agent-CLI 工具，并用 `myagents tool add` 注册进 MyAgents 工具注册表——注册后所有未来会话（builtin / DSH / Claude Code / Codex 全 runtime）的 AI 都会在 system prompt 里自动发现它。触发场景：(1) 用户说「把 XX 封装成工具」「做成一个工具」「注册个工具」「写个 CLI」「以后能直接用」；(2) 用户描述一个会反复出现的自动化需求——文档/文件批量处理、调用某个云 API、用某个多模态大模型做图像/视频理解等，即使没说"工具"两个字；(3) 你发现自己第二次为同类需求写几乎一样的脚本——这时要主动提议把它升格为注册工具，不要等用户开口。反向边界：一次性任务就地解决、不铸工具；接入现成的 MCP server 用 `myagents mcp`，不归这里。

Tool Creator — 创建并注册 Agent-CLI 工具

你正在 MyAgents 里运行。MyAgents 有一个工具注册表（~/.myagents/tools/）：注册进去的 CLI 工具会被投放到 PATH（~/.myagents/bin/），它的 description 会自动注入所有未来会话的 system prompt——未来的 AI（包括别的 runtime 上的）会自己想起它、查它的用法、调用它。用户也能在设置页「工具箱」里看到并管理它。

这个 skill 教你两件事：写出一个对 Agent 友好的合格 CLI 工具，以及把它注册进去。

第 0 步：判断要不要铸工具

铸一个工具是在为未来的几百次调用做投资，但注册表里的每个工具都占一行 system prompt。判断标准：

铸：需求会重复出现（用户明说"以后还要用"，或你已经第二次写同类脚本）；有清晰的输入→输出边界；参数可枚举。
不铸：一次性任务（就地写脚本跑完即弃）；纯交互探索类需求；已有注册工具能覆盖（先 myagents tool list 查一遍）。
模型管理器里已配置的模型、单发单收的调用——未来由 myagents model call 覆盖（若该命令存在，优先用它，不铸工具）。

灰色地带主动问用户："这个要不要我注册成工具，以后直接用？"

第 1 步：选原型

原型	特征	范例（动手前先读对应那个）
A · 纯本地处理	文件进文件出，零网络、零密钥	`references/example-local-tool.md`
B · API 包装	调外部 API，有密钥，可能多步工作流（上传→轮询→取结果）	`references/example-api-tool.md`

两个范例都是完整可跑的代码，照着改，不要从零发明结构。

第 2 步：写工具——形态契约

目录布局

~/.myagents/tools/<tool-name>/
├── tool.json      # manifest（注册时被读取校验）
└── run.mjs        # 入口，Node 单文件

tool.json

json

{
  "name": "md-merge",
  "version": "1.0.0",
  "description": "<≤800 字符，见第 3 步的撰写模板>",
  "entry": "run.mjs",
  "runtime": "node",
  "envKeys": [],
  "deps": []
}

name：kebab-case，3–30 字符。起名避开常见系统命令（curl、jq、git、node…）——~/.myagents/bin 在 PATH 里排在系统路径之前，重名会遮蔽系统命令，注册时会被直接打回。加领域前缀最稳妥（md-merge 而不是 merge）。
envKeys：工具需要的环境变量名列表（API key 等）。
deps：依赖的外部二进制（ffmpeg 等），没有就空数组。

技术栈

内置 Node v24 单文件 + node:util 的 parseArgs，零第三方依赖。这不是偏好是约束：MyAgents 已内置 Node 并打通全部 PATH，单文件意味着没有 node_modules、拷目录即分发、跨平台问题已被产品解决过一遍。确实绕不开外部二进制时声明进 deps 并做启动自检（见生死线 8）。

八条生死线

每条都解释了违反的后果——它们不是风格偏好，是工具在 Agent 手里能不能活的分界：

绝对禁止交互式输入。 任何 stdin prompt（确认、选择、密码）都会让 Agent 的 shell 调用永久挂死——Agent 没有键盘。危险操作（删除、覆盖、花钱）用确认协议替代：缺 --yes 时打印将要做的变更 + 完整的带 --yes 重跑命令，以退出码 4 退出。Agent 会把变更展示给用户、获准后重跑。
stdout 只放结果，stderr 放诊断，退出码语义化。 Agent 判断成败只靠这三样。退出码约定：0 成功 / 1 一般错误 / 2 用法或参数错误 / 3 环境缺失（缺 env key、缺依赖）/ 4 需要确认。进度提示、调试信息一律 stderr；spinner / 彩色转义码对 Agent 是纯噪音，不要。
必须有 --json 模式，错误也要结构化。 默认输出给人读的简洁文本；--json 输出机器可解析结果。出错时 stderr 给一行 JSON：{"error": "...", "code": "...", "remediation": "怎么修"}——remediation 是给下一个 Agent 的可行动建议（"run myagents tool env <name> set KEY=..."），不是模糊的 "something went wrong"。
密钥走 env，绝不走 argv。 argv 会泄进进程列表和日志。从 process.env[KEY] 读，缺失时按退出码 3 + remediation 处理。key 在 tool.json::envKeys 声明，由 myagents tool env <name> set KEY=value 设置。
有界运行时间。 不准变 daemon。所有网络请求带 AbortSignal.timeout(...)；轮询循环必须有次数/时间上限。预期超过 30 秒的操作在 readme 里声明耗时量级。
产物文件输出绝对路径到 stdout。 生成的文件落到当前工作区（或用户指定路径），把绝对路径作为结果打印——这是下游（预览、IM 发送）能接住产物的前提。
readme 子命令返回标准化使用文档。 固定章节（见第 3 步），未来的 AI 第一次用这个工具前会先跑 <tool> readme。
依赖自检，失败要可行动。 deps 里声明的二进制在启动时探测（等效 which），缺失则按生死线 3 的错误格式给出安装指引 + 退出码 3——Agent 拿到的是"装 ffmpeg：brew install ffmpeg"，不是一个看不懂的崩溃栈。

Show full SKILL.md (153 more words)Show less

三条进阶约定（让工具经得起时间）

接口即契约：工具被注册后，未来的 Agent 会从历史会话、readme 缓存里学到它的参数。改版时只加不改不删——加新参数可以，改既有参数的含义/删参数等于在所有学过它的 Agent 脚下抽地毯。
高信号输出：返回语义化字段（文件名、人类可读状态），不要裸 UUID / 内部 ID / 全量原始响应。输出大时做截断并提示如何取全量（--limit/--offset）。
本地先验证：参数格式、文件存在性在发网络请求之前验证，错误更快更准（退出码 2）。

第 3 步：写三面文档

工具有三个文档面，各答各的问题，不许混：

面	回答	消费者
`description`（tool.json）	"什么情况下该想起我"	system prompt 注入 + 设置页列表
`--help`	"参数怎么传"	AI 调用前现查 + 终端用户
`readme` 子命令	"解决这类问题的方法论"	AI 决定用之后 fetch + 设置页详情

description（≤800 字符，超长注册时被打回）

这 800 字符决定未来的 AI 会不会想起这个工具。按五件套写：

一句话能力声明
触发条件——给具体的用户说法（"用户要合并多个 markdown / 把文档拼成一个文件时"）。这是五件套里最值钱的部分：写"什么时候用"，不要重复"是什么"
反模式排除（如有）："单文件转换不要用这个，直接处理即可"
2–4 行 quick reference（最常用的调用形态）
收尾固定句式：首次使用前先运行 \<tool> readme`；机器可读输出加 --json`

写完做"新同事测试"：一个不知道这个工具存在的同事（或 AI），只读这 800 字符，能不能在对的时机想起它、并大致知道怎么开始？参数命名同理——--input-file 不要 --i，歧义是 Agent 的第一杀手。写不下的细节全部挪进 readme。

--help

必含：用法契约行（md-merge <files...> --out <path> [--json]）、每个参数一行说明、至少 3 条贴近真实的示例（Agent 靠示例适配自己的场景）、退出码表。

readme 子命令

固定章节，顺序不变（AI 每次读到的形状一致才能快速定位）：

markdown

# <tool-name>
## 何时使用       # 适用场景 + 不适用场景
## 快速开始       # 最小可用示例
## 参数           # 完整参数表（与 --help 一致）
## 示例           # 3+ 个真实场景的完整命令
## 失败模式       # 每种退出码/常见错误的含义与处理
## 产物回流       # 产物落在哪、怎么展示给用户（见下）

产物回流章节是强制的：写明"调用方（Agent）拿到产物路径后，必须在回复中引用它（Markdown 链接/图片）；IM 会话里用 myagents im send-media --file <path> 发送"。工具跑成功但用户看不到结果 = 这次调用白跑。

第 4 步：自测三连

注册前在 shell 里依次验证，任何一项不过就修：

bash

node run.mjs --help        # 用法契约行 + 示例齐全？
node run.mjs readme        # 六个固定章节齐全？
node run.mjs <真实参数>     # 一次真实调用：stdout 干净？exit code 对？--json 可解析？

原型 B 还要验：故意不设 env key 跑一次——应得到退出码 3 + 带 remediation 的 JSON 错误，而不是裸异常栈。

第 5 步：注册 + 告知

bash

myagents tool add ~/.myagents/tools/<tool-name>     # 校验 manifest、投 shim、进注册表
myagents tool env <tool-name> set API_KEY=<value>    # 原型 B：设密钥（让用户提供，绝不编造）
myagents tool list                                   # 确认出现在清单里

tool add 的常见打回：description 超 800 字符（精简后重试）、工具名撞系统命令（换名加前缀）。报错里带 recoveryHint，照做即可。

注册成功后必须在回复中明确告知用户（可审计性，不可省略）：

已注册工具 md-merge（合并多个 Markdown 文件）。我和之后的会话都能直接用它；你可以在设置 → 工具箱里查看、停用或删除它。

注册的工具对当前 session 之外的会话在它们下次启动时生效；你自己刚写完它，本 session 直接用就行。

如果 myagents tool --help 报 unknown command：当前 app 版本还没有注册机。把工具完整写好放在 ~/.myagents/tools/<name>/，告知用户"工具已就绪，等应用更新后运行 myagents tool add 注册"。

速查：完整流程

判断值不值得铸（第 0 步）
→ 选原型 A/B，读对应 references 范例（第 1 步）
→ 写 tool.json + run.mjs，过八条生死线（第 2 步）
→ 写 description 五件套 / --help / readme（第 3 步）
→ 自测三连（第 4 步）
→ tool add + env + 告知用户（第 5 步）

Files

SKILL.md and 2 other files (references) in bundled-skills/tool-creator of hAcKlyc/MyAgents.

SKILL.md
references/example-api-tool.md
references/example-local-tool.md

Open the folder on GitHubat commit f873ca3

Compare with similar skills

Tool Creator next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Tool Creator compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Tool Creator this skillhAcKlyc/MyAgents	915	—	~1.5k	Automated safety check: Pass	AGPL-3.0	today
NaturalNPC-Worldwide/npcpy	1.5k	—	~161	Automated safety check: Pass	MIT	yesterday
Pi AgentK-Dense-AI/scientific-agent-skills	48k	1 repos	~2.1k	Automated safety check: Pass	MIT	2 days ago
Workflow Schema Tuningbreaking-brake/cc-wf-studio	5.4k	—	~1.3k	Automated safety check: Pass	Custom licence	yesterday
Opikcomet-ml/opik-mcp	219	—	~2.1k	Automated safety check: Pass	Apache-2.0	yesterday
Add Promptcyanheads/pubmed-mcp-server	154	—	~1.6k	Automated safety check: Pass	Apache-2.0	3 days ago

Similar skills

Natural
NPC-Worldwide/npcpy
Render the provided prompt template with Jinja context and send it to the active NPC's LLM.
1.5k GitHub stars~161 tokensUpdated yesterday
AI & LLM EngineeringAuto-check passed
Pi Agent
K-Dense-AI/scientific-agent-skills
Builds with and operates Pi, the minimal terminal coding harness.
48k GitHub starsUsed in 1 repo~2.1k tokens
AI & LLM EngineeringAuto-check passed
Workflow Schema Tuning
breaking-brake/cc-wf-studio
Guides edits to cc-wf-studio's workflow schema so AI agents generate better workflows, treating schema text as prompt engineering rather than validation.
5.4k GitHub stars~1.3k tokensUpdated yesterday
Agent WorkflowsAuto-check passed
Opik
comet-ml/opik-mcp
Reference for the Opik SDK — tracing, span types, framework integrations, threads, and the prompt library (Python, TypeScript, REST).
219 GitHub stars~2.1k tokensUpdated yesterday
AI & LLM EngineeringAuto-check passed
Add Prompt
cyanheads/pubmed-mcp-server
Scaffold a new MCP prompt template. An agent skill from cyanheads/pubmed-mcp-server.
154 GitHub stars~1.6k tokensUpdated 3 days ago
Agent WorkflowsAuto-check passed
Opik Optimize
comet-ml/opik-mcp
Improve a prompt with the Opik Agent Optimizer — resolve the prompt, a dataset, and a metric, pick the algorithm, run a bounded optimization, check the gain on held-out data, and save the winner as…
219 GitHub stars~2.6k tokensUpdated yesterday
AI & LLM EngineeringAuto-check: notes

Questions about Tool Creator

What does Tool Creator do?

When should I use Tool Creator?

Tool Creator fits situations like: tasks that involve Prompt engineering; tasks that involve MCP servers.

How do I install Tool Creator in Claude Code?

Run `npx skills add hAcKlyc/MyAgents --skill tool-creator -a claude-code`. Or copy the skill folder (bundled-skills/tool-creator in hAcKlyc/MyAgents) into .claude/skills/tool-creator in your project. Claude Code loads it when a task matches its description.

How do I install Tool Creator in Codex?

Run `npx skills add hAcKlyc/MyAgents --skill tool-creator -a codex`. Or copy the skill folder (bundled-skills/tool-creator in hAcKlyc/MyAgents) into .agents/skills/tool-creator in your project. Codex loads it when a task matches its description.

Can I use Tool Creator in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add hAcKlyc/MyAgents --skill tool-creator -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/tool-creator, .gemini/skills/tool-creator, .github/skills/tool-creator and .opencode/skills/tool-creator in your project.

What does Tool Creator need to run?

Going by SKILL.md and its folder, Tool Creator needs the command-line tools its instructions call (node) and credentials named API_KEY. Our summary lists: A credential in API_KEY.

Does Tool Creator access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Tool Creator safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Tool Creator use?

Tool Creator is published under the AGPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Tool Creator use?

About 1.5k tokens (SKILL.md is roughly 6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 4k tokens, read only when the agent opens those files.

What are the alternatives to Tool Creator?

Skills that share tags, products or a category with Tool Creator: Natural (NPC-Worldwide/npcpy, 1.5k stars), Pi Agent (K-Dense-AI/scientific-agent-skills, 48k stars), Workflow Schema Tuning (breaking-brake/cc-wf-studio, 5.4k stars) and Opik (comet-ml/opik-mcp, 219 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Tool Creator?

hAcKlyc (a GitHub user) maintains it in hAcKlyc/MyAgents, which has 915 GitHub stars. The repository holds 15 skills in this directory. The repository was last updated on October 7, 2026.

Source: hAcKlyc/MyAgents on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install Tool Creator

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Context cost

Safety

SKILL.md

Tool Creator — 创建并注册 Agent-CLI 工具

第 0 步：判断要不要铸工具

第 1 步：选原型

第 2 步：写工具——形态契约

目录布局

tool.json

技术栈

八条生死线

三条进阶约定（让工具经得起时间）

第 3 步：写三面文档

description（≤800 字符，超长注册时被打回）

--help

readme 子命令

第 4 步：自测三连

第 5 步：注册 + 告知

速查：完整流程

Files

Compare with similar skills

Similar skills

More from hAcKlyc/MyAgents

Related

Questions about Tool Creator