Token导航 LogoToken导航

Agent Skill · 研究检索

role-creator(角色创造者)

role-creator 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

role-creator 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描有提醒项,按安装量在研究检索类的 73,675 个 Skill 里排第 40,402。

总安装 1,294周安装 55GitHub Stars 24下载 453核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:role-creator(角色创造者)
来源仓库:https://github.com/jsonlee12138/agent-team
仓库路径:skills/role-creator
安装命令:
npx skills add https://github.com/jsonlee12138/agent-team --skill role-creator
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/jsonlee12138/agent-team --skill role-creator

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

用于查找、检索和筛选相关信息。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。适合按关键词快速定位候选结果。可结合来源仓库 README 进一步核验具体用法。安装前建议确认是否会触发联网或文件操作。

role-creator 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/role-creator

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/jsonlee12138/agent-team --skill role-creator;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

role-creator 怎么安装?

在终端执行 npx skills add https://github.com/jsonlee12138/agent-team --skill role-creator。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

role-creator 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

role-creator 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

role-creator 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

role-creator 用的人多吗?

公开数据是累计安装 1,294 次,近一周安装 55 次,下载 453 次,GitHub 24 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

role-creator 安全吗?

第三方扫描有提醒项,说明存在需要人工确认的行为。安装前请打开源码看清它读了什么、联网到哪里。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方扫描存在提醒项或未通过项,需要人工逐条确认。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
role-creator

角色创造者 · 本页

1,29424需要联网未明确有提醒

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

role-creator

Generate a role skill package in a fixed contract:

  • <target>/SKILL.md
  • <target>/references/role.yaml
  • <target>/system.md

Target directory options:

  • skills/<role-name>/ — for open-source publishing (default)
  • .agent-team/teams/<role-name>/ — for team use with agent-team
  • <custom-path>/<role-name>/ — for custom local output paths

Required Workflow

  1. Normalize Input — validate role name as kebab-case.
  2. Generate Role Fields — auto-generate fields, delegate to /brainstorming on rejection.
  3. Select Skills — find-skills recommendations, user selection, manual additions.
  4. Execute CLI — run agent-team role create with approved parameters.
  5. Validate Output — verify three managed files match expected template structure.

Step 1: Normalize Input

  • role-name must be kebab-case. The current CLI requires it as a positional argument.
  • If the input is not kebab-case, suggest a normalized version and confirm with user.

Step 2: Generate Role Fields

  1. Auto-generate the following fields (do not ask user to draft them by default):

- description - system goal - in-scope (comma-separated) - out-of-scope (comma-separated)

  1. Present generated fields for user approval.
  2. If user rejects, or AI confidence is low (ambiguous scope, conflicting boundaries, vague goals), delegate to /brainstorming skill for structured refinement.
  3. After fields are approved, ask user for target directory (skills, .agent-team/teams, or a custom path).
  4. If target is skills or .agent-team/teams, note that the CLI checks matching global roles in ~/.agents/roles/ unless --force is passed. If matches are found, the CLI prompts whether to continue creating the local role.

Step 3: Select Skills

  1. Run find-skills to get recommendation candidates.
  2. Show recommendation list and let user select desired skills.
  3. Ask for manual additions after selection.
  4. When confirming the final skill identifiers, prefer the full remote identifier if the skill is known from a remote source (for example jsonlee12138/prompts@design-patterns-principles).
  5. Only use a local short name (for example design-patterns-principles) when no matching remote skill is found and the skill exists only locally.
  6. Merge selected + manual additions with de-duplication, then confirm final list.
  7. Final skills may be empty — will be persisted as skills: [] in references/role.yaml; otherwise persist skills[{name, description}] using the preferred identifier rule above.

If find-skills is unavailable or returns empty, skip recommendations and ask for manual additions only. In that fallback path, still prefer a full remote identifier if the user already provides one; otherwise use a local short name only for local-only skills.

Step 4: Execute CLI

CLI Reference

FlagTypeRequiredDefaultDescription
<role-name>argyes—Role name (kebab-case)
--descriptionstringyes—Role description
--system-goalstringyes—Primary objective for system.md
--in-scopestring[]no(from description)In-scope items (repeatable, comma-separated)
--out-of-scopestring[]nofallback textOut-of-scope items (repeatable, comma-separated)
--skillsstringno""Final selected skills (comma-separated)
--recommended-skillsstringno""Recommended skills from find-skills
--add-skillsstringno""Skills to add on top of selection
--remove-skillsstringno""Skills to remove from candidate list
--manual-skillsstringno""Manual fallback when recommendations unavailable
--target-dirstringnoskillsTarget: skills, .agent-team/teams, or custom path
--overwritestringnoaskOverwrite mode: ask/yes/no
--repo-rootstringno.Repository root path
--forceboolnofalseSkip global duplicate check

Skills resolution priority: --skills > --recommended-skills > --manual-skills, then --add-skills appended, --remove-skills excluded.

Examples

For open-source publishing (default):

agent-team role create frontend-dev \
  --description "Frontend role for UI implementation" \
  --system-goal "Ship accessible and maintainable UI work" \
  --in-scope "Build components,Improve accessibility" \
  --out-of-scope "Database migrations,Backend API ownership" \
  --skills "ui-ux-pro-max,better-icons"

For team use (agent-team integration):

agent-team role create frontend-dev \
  --target-dir .agent-team/teams \
  --description "Frontend role for UI implementation" \
  --system-goal "Ship accessible and maintainable UI work" \
  --in-scope "Build components,Improve accessibility" \
  --out-of-scope "Database migrations,Backend API ownership" \
  --skills "ui-ux-pro-max,better-icons"

Empty skills example:

agent-team role create product-manager \
  --description "Product role for roadmap and PRD work" \
  --system-goal "Define clear product requirements and priorities"

Step 5: Validate Output

Verify three files against actual template structure:

  1. SKILL.md — contains the role name, generated description frontmatter, and links to references/role.yaml and system.md.
  2. references/role.yaml — must contain:

- name — role name - description — role description - system_prompt_file: system.md - scope.in_scope — list of in-scope items - scope.out_of_scope — list of out-of-scope items - constraints.single_role_focus: true - skills — selected skills as objects with name and description (or empty [])

  1. system.md — contains system goal and operating constraints.

If any file is missing or contains unexpected content relative to the current templates, report the discrepancy and offer to regenerate.

Overwrite Behavior

  • Controlled by --overwrite flag (ask/yes/no).
  • Current CLI does not create a backup directory when overwriting.
  • Only managed files are overwritten (SKILL.md, references/role.yaml, system.md); other files in the role directory are preserved.
  • Legacy root-level role.yaml is removed if present after generation.

Runtime Skill Discovery (Role Usage Guideline)

Generated roles should follow this behavior at runtime:

  1. When a role receives a task that requires a skill not listed in its references/role.yaml, it should invoke find-skills to search for a matching skill.
  2. If a suitable skill is found, it may only attempt project-level installation or use an already available local/project-cached skill.
  3. Global installation is not allowed for worker/runtime resolution.
  4. If the skill is still unavailable locally, emit a warning with the reason and suggested next step, then continue with best-effort execution.
  5. After successful use, suggest adding the skill to the role's references/role.yaml for future sessions by regenerating the role.
  6. If find-skills is unavailable or returns no match, the role should notify the user and proceed with best-effort execution.

平台分布

Codex

35.76%
按下载量换算162

Claude

31.71%
按下载量换算144

Cursor

18.59%
按下载量换算84

Gemini CLI

10.06%
按下载量换算46

安全审计

Gen Agent Trust Hub

可疑

Socket

可疑

Snyk

可疑

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在提醒或未通过项,安装前请自行复核。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills