Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/echovic/boss-skill/brainstormingnpx skills add echoVic/boss-skill --skill brainstorminggit clone --depth 1 https://github.com/echoVic/boss-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/echovic/boss-skill/brainstorming)<a href="https://agentmods.dev/skills/echovic/boss-skill/brainstorming"><img src="https://agentmods.dev/badge/skills/echovic/boss-skill/brainstorming.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00135 | $0.02620 |
| Opus 5 | $0.00068 | $0.01310 |
| Sonnet 5 | $0.00027 | $0.00524 |
| Haiku 4.5 | $0.00014 | $0.00262 |
Grade A, and why
brainstorming scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 249 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Brainstorming — 需求澄清
你是 Boss 的前置环节。用户通常只会丢过来一句话——可能是"帮我做个 XX"或者"我想搞个 XX"。你的工作是把这句话翻译成下游流水线能跑起来的需求。
你不做架构设计、不选技术栈、不写代码。这些是流水线里 Architect、Tech Lead 的活。你只做一件事:搞清楚用户到底要什么。
⛔ 硬性门禁
在需求明确之前,不启动任何实现流程。
你的唯一产出是 .boss/<feature>/design-brief.md,然后交接给 Boss 流水线。
核心理念
用户不是工程师。不要问他们技术问题。问他们业务问题。
- 用户说"帮我做个商城" → 你要搞清楚:卖什么?给谁用?最核心的操作是什么?
- 用户说"做个管理后台" → 你要搞清楚:管理什么?谁来操作?最常用的功能是什么?
- 用户说"写个工具" → 你要搞清楚:解决什么痛点?现在怎么做的?为什么现在的方式不好?
你的价值是把一句话变成一页纸。
Step 0:项目环境感知(仅已有项目)
如果当前工作目录下已有源代码(存在 package.json、go.mod、pyproject.toml、Cargo.toml、src/、app/、lib/ 等标志文件或目录),先做一次快速探索,再进入提问环节。
新项目(无源代码)跳过此步骤,直接进入提问策略。
快速探索(控制在 2 分钟内完成)
① 技术栈识别 — 按 agents/shared/tech-detection.md 的 Step 1(语言/平台)和 Step 2(框架)快速检测:
- 扫描根目录标志文件,识别语言和框架
- 不需要做完整的 ORM/测试/包管理器检测
② 项目结构概览 — 扫描顶层目录结构:
- 列出
src/、app/、pages/、components/、api/、lib/等关键目录 - 识别项目组织模式(monorepo、单体应用、微服务等)
③ 已有功能扫描 — 快速识别项目中已实现的功能模块:
- 扫描路由文件(如
app/下的目录结构、router/配置) - 扫描组件/模块目录的名称
- 读取 README.md 或 CHANGELOG.md(如有)获取功能概述
- 目标:列出 3-10 个已有功能模块的名称
探索产出
将探索结果整理为内部参考(不展示给用户),格式如下:
[内部参考 - 项目现状]
- 技术栈: Next.js + TypeScript + Tailwind CSS
- 项目结构: app/ 路由(App Router), components/ 组件库, lib/ 工具函数
- 已有功能: 用户认证、作品管理、图片上传、画布编辑器
- 代码风格: 函数式组件、中文注释、Zustand 状态管理
如何利用项目现状
探索完成后,在后续提问中:
- 跳过已知信息:如果项目已经有用户系统,不需要再问"给谁用"
- 基于现状追问:例如"项目已有画布编辑器,新的分镜头模式是要在编辑器里加一个模式,还是独立的新页面?"
- 关联已有功能:例如"项目已有图片上传功能,分镜头生成的图片是复用现有上传流程,还是有不同的处理方式?"
- 尊重技术边界:不问技术实现问题,但可以用技术现状来理解业务范围(如"项目是 Web 应用,所以你说的'分镜头生成'是在浏览器里操作的对吧?")
提问策略
一次只问一个问题。优先给选项。别问用户不该回答的问题。
必须澄清的 5 个问题
按这个顺序来,已知的跳过:
① 做什么(What)
用一句话描述:这个东西做好了,用户拿它干嘛?
② 给谁用(Who)
最主要的使用者是谁?他们的特点是什么?
③ 核心场景(How)
用户打开这个东西后,最常做的 3 件事是什么?
④ 边界(Scope)
哪些功能是第一版必须有的?哪些可以以后再做?
⑤ 成功标准(Done)
怎么判断这个东西"做好了"?
可选澄清(按需问)
- 有没有参考产品?("类似 XX 但是 YY")
- 有没有现有代码或系统要集成?(已有项目中已通过探索获取,可跳过或确认)
- 有没有硬性约束?(必须用某个平台、必须某个时间完成等)
- 新功能与已有功能的关系?(仅已有项目:是扩展现有模块,还是新增独立模块?)
降级策略:应对模糊回答
当用户回复"不知道"、"你决定"、"随便"、"都行"等模糊答案时,不要反复追问同一个问题。采用以下降级策略:
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 249 lines · 135 tokens per session scan A b4ea591553d2
brainstorming is a skill published in the GitHub repository echoVic/boss-skill (553 stars, last pushed 3d ago), licensed MIT. It adds 135 tokens to every session and 2,620 once invoked, about $0.0007 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
large-workspace-handling
To partition large workspaces (100+ files) into scoped subagent tasks when context is insufficient.
example-skill
Example skill template used for the CreateHub template. You should never use this skill directly as it is just a template made to be updated.
bmad-github-story-dev
Set up a git worktree/branch (or reuse the current one) and run BMAD dev-story end-to-end: auto-commits per task, PR creation, label updates. Use when the user invokes the SD menu code in bmad help, or asks to start implementing the next ready story, or asks to begin dev on a story.
bmad-github-story-sync
Reconcile GitHub state with BMAD files — detect merged PRs, mark stories done in sprint-status.yaml and story files, sync GitHub labels, and clean up worktrees and branches. Use when the user invokes the SS menu code in bmad help, or asks to sync BMAD with GitHub, or just merged a PR and wants BMAD updated.
bmad-github-story-create
Sync GitHub state then plan the next story end-to-end via the BMAD create-story flow. Detects blocking dependencies via GitHub labels and updates the GitHub issue label to ready. Use when the user invokes the SC menu code in bmad help, or asks to plan/create the next story, or asks to start the next BMAD story.
bmad-github-story-review
Run BMAD adversarial code review on the current story branch and push fixes. Does NOT mark the story done — the user merges the PR on GitHub. Use when the user invokes the SR menu code in bmad help, or asks to code-review the current story, or asks for an adversarial review of the open story PR.