PRODUCT.md 生成工具 - 交互式产品文档创建，解锁智能委员会评审

product by boshu2/agentops

261 周安装量

222 GitHub Stars

GitHub

安装命令

npx skills add https://github.com/boshu2/agentops --skill product

自动化技术文档产品管理

🇨🇳中文介绍

/product — 交互式 PRODUCT.md 生成

目的： 引导用户创建 PRODUCT.md，以解锁 /pre-mortem 和 /vibe 中基于产品认知的委员会评审。

你必须执行此工作流。不要仅仅描述它。

CLI 依赖项： 无要求。

执行步骤

给定 /product [target-dir]：

target-dir 默认为当前工作目录。

步骤 1：预检

检查 PRODUCT.md 是否已存在：

ls PRODUCT.md 2>/dev/null

如果存在：

使用 AskUserQuestion：

广告位招租

在这里展示您的产品或服务

触达数万 AI 开发者，精准高效

联系我们

步骤 2：收集上下文

读取可用的项目文件以预填充建议：

README.md — 提取项目描述、目的、目标受众
package.json / pyproject.toml / go.mod / Cargo.toml — 提取项目名称
目录列表 — 对项目根目录执行 ls 以获取结构提示

利用你的发现为每个部分起草初始建议。如果没有任何文件存在，则使用空白建议继续。

使用 AskUserQuestion 向用户询问每个部分。对于每个问题，尽可能提供步骤 2 中获得的预填充建议。

询问："您产品的使命是什么？（一句话：它为谁做什么？）"

基于 README 分析的选项：

从 README 推导出的建议使命（如果可用）
一个更简短/更有力的变体
"让我自己输入"

3b：目标用户画像

询问："您的主要用户是谁？描述 2-3 个用户画像。"

为每个画像收集：

角色（例如，"后端开发人员"、"DevOps 工程师"）
目标 — 他们试图完成什么
痛点 — 当前是什么让这变得困难

使用 AskUserQuestion 询问第一个画像的角色，然后以对话方式跟进细节和额外的画像。当用户表示完成或收集了 3 个画像后停止。

3c：核心价值主张

询问："是什么让您的产品值得使用？列出 2-4 个关键价值主张。"

从 README/项目上下文推导出的建议
"让我自己输入"

询问："存在哪些替代方案，您如何区分？"

替代方案名称
它们的优势
您的差异化点

如果用户说"没有"或"跳过"，则在该部分写入"未识别出直接竞争对手"。

步骤 4：生成 PRODUCT.md

将 PRODUCT.md 写入目标目录，结构如下：

---
last_reviewed: YYYY-MM-DD
---

# PRODUCT.md

## 使命

{来自 3a 的使命}

## 目标用户画像

### 画像 1: {角色}
- **目标:** {目标}
- **痛点:** {痛点}

{为每个画像重复}

## 核心价值主张

{来自 3c 的要点列表}

## 竞争格局

| 替代方案 | 优势 | 我们的差异化点 |
|-------------|----------|---------------------|
{来自 3d 的行}

## 用法

此文件启用基于产品认知的委员会评审：

- **`/pre-mortem`** — 当此文件存在时，自动包含 `product` 视角（用户价值、采用障碍、竞争地位）以及计划评审评委。
- **`/vibe`** — 当此文件存在时，自动包含 `developer-experience` 视角（API 清晰度、错误体验、可发现性）以及代码评审评委。
- **`/council --preset=product`** — 按需运行产品评审。
- **`/council --preset=developer-experience`** — 按需运行 DX 评审。

用户显式指定的 `--preset` 覆盖会跳过自动包含（用户意图优先）。

将 last_reviewed 设置为今天的日期（YYYY-MM-DD 格式）。

创建了什么： {path} 处的 PRODUCT.md
它解锁了什么：
- /pre-mortem 现在将自动包含产品视角（用户价值、采用障碍、竞争地位）
- /vibe 现在将自动包含开发者体验视角（API 清晰度、错误体验、可发现性）
- /council --preset=product 和 /council --preset=developer-experience 可按需使用
后续步骤： 建议在他们下一个计划上运行 /pre-mortem 以查看产品视角的实际效果

为新项目创建产品文档

用户说： /product

发生的情况：

代理检查是否存在 PRODUCT.md，未找到
代理读取 README.md 和 package.json 以提取项目上下文
代理询问用户使命，建议"用于自动化依赖项更新的 CLI 工具"
代理访谈 2 个画像：DevOps 工程师和 Backend Developer
代理询问价值主张，用户提供："零配置自动化、安全更新、节省时间"
代理询问竞争对手，用户提到 Renovate 和 Dependabot
代理使用所有收集的信息和今天的评审日期写入 PRODUCT.md

结果： 创建了 PRODUCT.md，在未来的验证中解锁了基于产品认知的委员会视角。

更新现有产品文档

用户说： /product

发生的情况：

代理找到 3 个月前存在的 PRODUCT.md
代理提示："PRODUCT.md 已存在。您希望如何操作？"
用户选择"更新 — 保留现有内容作为默认值"
代理读取当前文件，提取使命和画像作为建议
代理询问使命，用户保留现有的
代理询问画像，用户添加新的"安全工程师"画像
代理使用新画像更新 PRODUCT.md，更新 last_reviewed 日期

结果： PRODUCT.md 已刷新，添加了新画像，为下一个验证周期做好准备。

问题	原因	解决方案
没有上下文来预填充建议	缺少 README 或项目元数据文件	使用空白建议继续。请用户用自己的话描述项目。从对话中提取使命。
用户不清楚画像与用户的区别	对画像定义感到困惑	解释："画像是具有目标和痛点的特定用户原型。想象一个会使用此产品的真实人物。"提供示例。
竞争格局感觉牵强	真正新颖的产品或小众工具	接受"无直接竞争对手"为有效答案。专注于替代方法（手动流程、脚本）而非产品。
PRODUCT.md 感觉泛泛而谈	用户输入不足或访谈仓促	提出后续问题。请求具体示例。挑战模糊的陈述，如"让事情更容易" — 如何更容易？如何衡量？

🇺🇸English

/product — Interactive PRODUCT.md Generation

Purpose: Guide the user through creating a PRODUCT.md that unlocks product-aware council reviews in /pre-mortem and /vibe.

YOU MUST EXECUTE THIS WORKFLOW. Do not just describe it.

CLI dependencies: None required.

Execution Steps

Given /product [target-dir]:

target-dir defaults to the current working directory.

Step 1: Pre-flight

Check if PRODUCT.md already exists:

ls PRODUCT.md 2>/dev/null

If it exists:

Use AskUserQuestion:

Question: "PRODUCT.md already exists. What would you like to do?"
Options:
- "Overwrite — start fresh" → continue to Step 2
- "Update — keep existing content as defaults" → read existing file, use its values as pre-populated suggestions in Step 3
- "Cancel" → stop, report no changes

If it does not exist: continue to Step 2.

Step 2: Gather Context

Read available project files to pre-populate suggestions:

README.md — extract project description, purpose, target audience
package.json / pyproject.toml / go.mod / Cargo.toml — extract project name
Directory listing — ls the project root for structural hints

Use what you find to draft initial suggestions for each section. If no files exist, proceed with blank suggestions.

Step 3: Interview

Ask the user about each section using AskUserQuestion. For each question, offer pre-populated suggestions from Step 2 where available.

3a: Mission

Ask: "What is your product's mission? (One sentence: what does it do and for whom?)"

Options based on README analysis:

Suggested mission derived from README (if available)
A shorter/punchier variant
"Let me type my own"

3b: Target Personas

Ask: "Who are your primary users? Describe 2-3 personas."

For each persona, gather:

Role (e.g., "Backend Developer", "DevOps Engineer")
Goal — what they're trying to accomplish
Pain point — what makes this hard today

Use AskUserQuestion for the first persona's role, then follow up conversationally for details and additional personas. Stop when the user says they're done or after 3 personas.

3c: Core Value Propositions

Ask: "What makes your product worth using? List 2-4 key value propositions."

Options:

Suggestions derived from README/project context
"Let me type my own"

3d: Competitive Landscape

Ask: "What alternatives exist, and how do you differentiate?"

Gather:

Alternative names
Their strengths
Your differentiation

If the user says "none" or "skip", write "No direct competitors identified" in the section.

Step 4: Generate PRODUCT.md

Write PRODUCT.md to the target directory with this structure:

---
last_reviewed: YYYY-MM-DD
---

# PRODUCT.md

## Mission

{mission from 3a}

## Target Personas

### Persona 1: {role}
- **Goal:** {goal}
- **Pain point:** {pain point}

{repeat for each persona}

## Core Value Propositions

{bullet list from 3c}

## Competitive Landscape

| Alternative | Strength | Our Differentiation |
|-------------|----------|---------------------|
{rows from 3d}

## Usage

This file enables product-aware council reviews:

- **`/pre-mortem`** — Automatically includes `product` perspectives (user-value, adoption-barriers, competitive-position) alongside plan-review judges when this file exists.
- **`/vibe`** — Automatically includes `developer-experience` perspectives (api-clarity, error-experience, discoverability) alongside code-review judges when this file exists.
- **`/council --preset=product`** — Run product review on demand.
- **`/council --preset=developer-experience`** — Run DX review on demand.

Explicit `--preset` overrides from the user skip auto-include (user intent takes precedence).

Set last_reviewed to today's date (YYYY-MM-DD format).

Step 5: Report

Tell the user:

What was created: PRODUCT.md at {path}
What it unlocks:
- /pre-mortem will now auto-include product perspectives (user-value, adoption-barriers, competitive-position)
- /vibe will now auto-include developer-experience perspectives (api-clarity, error-experience, discoverability)
- /council --preset=product and /council --preset=developer-experience are available on demand
Next steps: Suggest running /pre-mortem on their next plan to see product perspectives in action

Examples

Creating Product Doc for New Project

User says: /product

What happens:

Agent checks for existing PRODUCT.md, finds none
Agent reads README.md and package.json to extract project context
Agent asks user about mission, suggesting "CLI tool for automated dependency updates"
Agent interviews for 2 personas: DevOps Engineer and Backend Developer
Agent asks about value props, user provides: "Zero-config automation, Safe updates, Time savings"
Agent asks about competitors, user mentions Renovate and Dependabot
Agent writes PRODUCT.md with all gathered information and today's review date

Result: PRODUCT.md created, unlocking product-aware council perspectives in future validations.

Updating Existing Product Doc

User says: /product

What happens:

Agent finds existing PRODUCT.md from 3 months ago
Agent prompts: "PRODUCT.md exists. What would you like to do?"
User selects "Update — keep existing content as defaults"
Agent reads current file, extracts mission and personas as suggestions
Agent asks about mission, user keeps existing one
Agent asks about personas, user adds new "Security Engineer" persona
Agent updates PRODUCT.md with new persona, updates last_reviewed date

Result: PRODUCT.md refreshed with additional persona, ready for next validation cycle.

Troubleshooting

Problem	Cause	Solution
No context to pre-populate suggestions	Missing README or project metadata files	Continue with blank suggestions. Ask user to describe project in own words. Extract mission from conversation.
User unclear on personas vs users	Confusion about persona definition	Explain: "Personas are specific user archetypes with goals and pain points. Think of one real person who would use this." Provide example.
Competitive landscape feels forced	Genuinely novel product or niche tool	Accept "No direct competitors" as valid. Focus on alternative approaches (manual processes, scripts) rather than products.
PRODUCT.md feels generic	Insufficient user input or rushed interview	Ask follow-up questions. Request specific examples. Challenge vague statements like "makes things easier" — easier how? Measured how?

Weekly Installs

200

Repository

boshu2/agentops

GitHub Stars

197

First Seen

Feb 12, 2026

Security Audits

Gen Agent Trust HubPass SocketPass SnykPass

Installed on

opencode199

codex196

github-copilot195

gemini-cli195

kimi-cli193

amp193

通过 LiteLLM 代理让 Claude Code 对接 GitHub Copilot 运行 | 高级变通方案指南

27,600 周安装