What is product-architecture?

A product-analysis method that reverse-engineers how a product is organised, rather than mapping its technical services or page structure. It represents the product with four layers—touchpoints, user situations, capabilities, and data—and produces an HTML architecture diagram.

Use it to analyse an AI app or other product, group its visible features into business modules, infer shared capabilities and data sources, and create a visual architecture map.

It gives product teams a consistent way to understand what a product does, how its parts depend on one another, and how data moves through it. The method also helps compare products or find missing opportunities.

product-architecture — Skill for Claude Code and Codex by echoyu1025-a

Install

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.

Any agent

npx skills add echoyu1025-a11y/ai-product-skills --skill product-architecture

Clone the repo

git clone --depth 1 https://github.com/echoyu1025-a11y/ai-product-skills

Made for: Claude Code, Codex.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/echoyu1025-a11y/ai-product-skills/product-architecture/github.svg)](https://agentmods.dev/skills/echoyu1025-a11y/ai-product-skills/product-architecture)

Your own site

<a href="https://agentmods.dev/skills/echoyu1025-a11y/ai-product-skills/product-architecture"><img src="https://agentmods.dev/badge/skills/echoyu1025-a11y/ai-product-skills/product-architecture/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/echoyu1025-a11y/ai-product-skills/product-architecture"><img src="https://agentmods.dev/badge/skills/echoyu1025-a11y/ai-product-skills/product-architecture.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 167 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,663 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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 5.1	$0.00167	$0.01663
Opus 5	$0.00084	$0.00831
Sonnet 5	$0.00033	$0.00333
Haiku 4.5	$0.00017	$0.00166

Measured 12d ago against content hash 16713b9c8b9c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

product-architecture 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 12d 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.

product-architecture/SKILL.md · 104 lines

How it starts

The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.

产品架构 SKILL — 逆向拆解任意产品

这个 SKILL 是什么

把任意产品(微信、豆包、某 AI App、ToB 工具…)逆向拆解成产品经理视角的架构图:

四层骨架:触达层 / 场景层 / 能力层 / 数据层
数据流转:跨层箭头,标注沉淀和回流路径
可视化输出:HTML 文件,黄绿蓝紫四色分层 + 蓝/红双向箭头

不是技术架构(微服务、数据库选型),不是信息架构(页面结构),是产品架构 — 业务模块怎么切、模块之间什么关系、数据怎么流转。

触发后的工作流程

第 0 步:确认目标产品

如果用户没说清是哪个产品,先问:

你要拆解哪个产品?(名称 + 官网/截图/简介任一即可)
用户群是 ToC 还是 ToB?
你的拆解目的是:深度研究(看懂别人设计意图)还是 找空白点(找差异化机会)?

如果产品比较小众或用户没给信息,主动用 WebFetch / WebSearch 查官网了解。

第 1 步:阅读方法论

必读 references/methodology.md,这是整个 SKILL 的方法论核心。里面有:

四层骨架的层定义(触达/场景/能力/数据各是什么)
六步法(体验→穷举→归类→反推能力→反推数据→绘图)
粒度判断("能描述这功能帮用户做什么"就够了)
归类判断("砍掉 A 后 B 是否受影响"测试)

读完再继续。

第 2 步:参考案例

读 references/examples.md,里面有四个范例:

微信(传统 ToC)
豆包(AI ToC)
飞书 AI(AI ToB)
AI 搜索(Agent 平台)

挑一个跟目标产品最相似的形态重点看,理解"该层应该写什么粒度的东西"。

第 3 步:六步法逆向推导

按顺序执行,不能跳步:

体验:走一遍核心链路。能上手就上手,不能就看演示视频/官网截图。把核心场景在脑子里跑一遍。
穷举:列出所有用户可见的功能。颗粒度按对话类/输入类/生成类/工具类拆分。漏一个可能漏一整个业务模块。
归类:用"砍掉 A,B 是否受影响"测试,把功能合并成业务模块。形成 3-6 个场景层模块。
反推能力:每个场景层模块依赖什么 AI/通用能力?(大模型对话、RAG、ASR/TTS、图像生成、工作流引擎…)同一能力被多个场景共用时只画一次。
反推数据:每个能力消费什么数据?(企业存量文档、用户对话日志、互联网爬取、UGC、合成数据…)
绘图:用 templates/architecture-template.html 生成可视化。

第 4 步:识别数据流转

四层骨架只是静态结构,箭头才是产品的"活"。至少标 1-2 条:

向下(蓝色 ↓):用户行为沉淀到数据层。例:对话记录 → 沉淀到数据层用于训练
向上(红色 ↑):数据回流被产品消费。例:UGC 智能体 → 回流到场景层广场
横向:同层模块之间的依赖(罕见,慎用)

第 5 步:生成输出

复制 templates/architecture-template.html → 当前工作目录,文件名 <产品名拼音或英文>-architecture.html。

按 assets/color-tokens.md 的配色规范替换内容。不要改变四层颜色对应(黄=触达/绿=场景/蓝=能力/紫=数据),保持视觉一致性。

完成后告诉用户:

文件路径
一句话总结这个产品的架构特点(例如:"豆包是典型的 AI ToC 五层结构,生态层是它的增长引擎")
1-2 条值得注意的设计决策或空白点

关键原则

拒绝凑数。如果某层只有 1 个东西,要么是漏穷举了,要么是这个产品本身就薄 — 后一种情况要明确告诉用户。
不替代用户判断。架构是判断题不是计算题,同一个产品不同人画不同。输出后引导用户讨论:"你觉得我把 X 放在场景层还是能力层更合理?"
不画 UI。如果你忍不住写"登录页/设置页/个人中心",停下,这是信息架构。问自己:"这个东西帮用户做什么业务?"
保持视觉一致。配色规范已固定,不要自创色系。

常见误判

现象	误判	纠正
把"语音输入"放在场景层	❌	它是能力层,场景层是"语音对话/语音笔记"这种业务
把"用户画像"放在数据层	⚠️ 看情况	如果是用于推荐的特征工程产物,属于数据层;如果是 ToB 客户管理界面,属于场景层
触达层只写"App"	❌ 太粗	应该写"iOS App / Android App / Web 端 / 小程序 / 公众号"
数据层写"MySQL"	❌ 技术细节	写"用户对话记录 / 企业存量文档 / 互联网爬取"

Read the full file on GitHub · 104 lines

Files

What ships with it

4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

12d ago First seen · 104 lines · 167 tokens per session scan A 16713b9c8b9c

Subscribe to this mod's changes

product-architecture is a skill published in the GitHub repository echoyu1025-a11y/ai-product-skills (3 stars, last pushed 2mo ago), licensed MIT. It adds 167 tokens to every session and 1,663 once invoked, about $0.0008 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 2026-08-31.

Claude Code Design and UX Planning agent-skills ai-product-manager chinese claude-skills prd

product-architecture