Token导航 LogoToken导航

Agent Skill · 研究检索

bmad-idea(疯狂的想法)

bmad-idea 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。

bmad-idea 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 64,068。

总安装 306周安装 13GitHub Stars 2下载 107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:bmad-idea(疯狂的想法)
来源仓库:https://github.com/akillness/oh-my-gods
仓库路径:skills/bmad-idea
安装命令:
npx skills add https://github.com/akillness/oh-my-gods --skill bmad-idea
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/akillness/oh-my-gods --skill bmad-idea

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。

用于创意构思和问题重构,适合 PRD 编写前的概念探索阶段。适用于头脑风暴、JTBD 框架分析或商业模式选项评估等前期发散任务。使用时需明确创意筛选标准,避免陷入无限发散而缺乏落地路径。

安装命令:npx skills add https://github.com/akillness/oh-my-gods --skill bmad-idea。建议配合用户画像和竞品分析使用,确保想法具备可行性而非纯幻想。

来源仓库内路径:skills/bmad-idea

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/akillness/oh-my-gods --skill bmad-idea;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

bmad-idea 怎么安装?

在终端执行 npx skills add https://github.com/akillness/oh-my-gods --skill bmad-idea。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

bmad-idea 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。

bmad-idea 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。

bmad-idea 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。

bmad-idea 用的人多吗?

公开数据是累计安装 306 次,近一周安装 13 次,下载 107 次,GitHub 2 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。

bmad-idea 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
bmad-idea

疯狂的想法 · 本页

3062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

bmad-idea

Use bmad-idea when the job is shaping an idea before the team commits to PRDs, architecture, or implementation. Keep the entrypoint focused on choosing the right creative lane and the next command. Push command catalogs, persona detail, and method inventories into supporting references.

When to use this skill

  • Brainstorm new concepts, variations, or option sets before narrowing scope
  • Run design thinking for user-centered reframing, empathy work, or prototype direction
  • Explore innovation strategy, differentiation, JTBD framing, or business-model options
  • Diagnose a messy problem before roadmap or implementation planning
  • Turn a concept into a narrative, pitch, or product story
  • Combine multiple creative lanes only when the user truly needs a broader "creative squad" pass

Use a different skill instead when:

  • the user already needs phased delivery, artifact progression, or BMAD status: use bmad
  • the user wants a specific doc, backlog, or implementation plan more than ideation: use a narrower planning or writing skill
  • the user is already executing work across agents: use an execution workflow such as omg, omc, omx, or team

Instructions

Step 1: Classify the creative job before naming commands

Sort the request into one primary lane:

  • brainstorming: widen the option space, generate directions, remix concepts
  • design-thinking: center users, workflows, empathy, prototypes, or testing
  • innovation-strategy: identify market gaps, moats, positioning, or business models
  • problem-solving: diagnose root causes, system constraints, or repeated blockers
  • storytelling: package the idea into a pitch, narrative, or messaging arc
  • multi-lane: use only when the request clearly needs several of the above in one pass

Do not dump all CIS workflows before the main lane is clear.

Step 2: Choose the lightest workflow that fits

Route directly:

  • brainstorming -> bmad-cis-brainstorming or /cis-brainstorm
  • design-thinking -> bmad-cis-design-thinking or /cis-design-thinking
  • innovation-strategy -> bmad-cis-innovation-strategy or /cis-innovation-strategy
  • problem-solving -> bmad-cis-problem-solving or /cis-problem-solving
  • storytelling -> bmad-cis-storytelling or /cis-storytelling
  • multi-lane -> creative squad

If the user wants to keep talking to one specialist rather than run the whole workflow, load the matching /cis-agent-* surface from references/workflow-map.md.

Step 3: Load only the supporting detail you need

Use progressive disclosure:

  • references/workflow-map.md for commands, direct agent loading, output shapes, and when to use creative squad
  • references/agent-profiles.md for persona fit, methods, and visual-tool notes when those details actually matter to the response

Keep the main skill compact instead of re-expanding every workflow catalog inline.

Step 4: Preserve the boundary between ideation and delivery

Use bmad-idea to shape what should happen next, not to force structured delivery too early:

  • if the user is still forming the concept, stay here
  • if the user now needs PRDs, architecture, sprint planning, or workflow status, route back to bmad
  • if the user needs implementation execution, hand off to the appropriate execution workflow rather than keeping everything in CIS

Step 5: Verify the creative lane actually fits

Before claiming success, prove that:

  • the chosen lane matches the user's immediate job
  • the response names the relevant command or specialist instead of the whole suite
  • the expected output shape is clear enough for the next step
  • the answer did not accidentally collapse back into generic BMAD phase routing

Examples

Example 1: Brainstorm before specs

Input:

Brainstorm three directions for a habit-tracking app before we commit to specs.

Expected shape:

  • treats this as ideation rather than structured delivery
  • routes to brainstorming instead of the full BMAD phase path
  • names Carson or the brainstorming command as the next move

Example 2: Use design thinking for a user flow

Input:

We need empathy mapping and quick prototype directions for our onboarding flow.

Expected shape:

  • identifies this as design-thinking work
  • routes to Maya or the design-thinking command
  • stays focused on user framing instead of dumping the whole CIS catalog

Example 3: Diagnose before roadmapping

Input:

Players love the core loop but churn after day 2. Diagnose the problem before we write a roadmap.

Expected shape:

  • treats the request as diagnosis and constraint discovery
  • routes to Dr. Quinn or the problem-solving workflow
  • avoids jumping straight to solutioning or roadmap planning

Example 4: Hand off once ideation is done

Input:

We already shaped the idea. Which BMAD phase should come next?

Expected shape:

  • recognizes that the job has moved from ideation into structured delivery
  • routes the request back to bmad
  • avoids keeping the user inside the creative suite unnecessarily

Best practices

  1. Pick one primary creative lane before listing commands
  2. Default to the narrowest useful workflow, not creative squad
  3. Keep ideation work distinct from BMAD phase routing and implementation
  4. Load agent profiles or command maps only when the extra detail is needed
  5. Prefer user-facing outcomes over method-dump answers
  6. Route back to bmad once the idea is ready for structured delivery
  7. Keep the entrypoint compact and move catalog detail into references

References

平台分布

Codex

38.06%
按下载量换算41

Claude

28.59%
按下载量换算31

Cursor

19.42%
按下载量换算21

Gemini CLI

9.37%
按下载量换算10

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

Snyk

通过

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