Remote OpenClaw
Menu
SkillsMCPPluginsFree guideDigestSubmit MCPSkillPluginMCPMCP, plugin, or skillAdvertise
Remote OpenClaw
SkillsMCPPluginsFree guideDigestSubmit MCPSkillPluginMCPMCP, plugin, or skillAdvertise

Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →
Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →
Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →
Skills/liuzhengdongfortest/codestable/cs-roadmap
cs-roadmap logo

cs-roadmap

liuzhengdongfortest/codestable
998 installs923 stars
Run it on Hostinger →up to 70% off + an extra 10% with code ZACAARON10Free API →

Installation

npx skills add https://github.com/liuzhengdongfortest/codestable --skill cs-roadmap

Summary

把"大到塞不进单个 feature"的需求做成完整事前规划:概设 + 接口契约 + 子 feature 拆解清单,放在 `.codestable/roadmap/{slug}/`。两种模式 new / update。触发:用户说"我想要一个 X 系统"、"帮我把这块需求拆一下"、"开一份 roadmap",或 feature-design 阶段发现需求太大。

SKILL.md

cs-roadmap

启动必读

开始任何判断或动作前,先读取 .codestable/attention.md;缺失则视为骨架不完整,提示先补齐或运行 cs-onboard,不要回退到外部 AI 入口文件。

.codestable/roadmap/ 是项目的"规划层"——每个子目录承载一块大需求,主文档由三块构成:

  1. 概设:这块大需求要怎么搭、拆成哪几个模块 / 组件、各自职责
  2. 架构层详设:模块之间的接口契约、共享数据结构、跨 feature 的协议
  3. 子 feature 拆解:把方案分解成一串带依赖关系的子 feature 种子,feature 流程一次消费一条

三块一起作为这块大需求所有子 feature 的共同约束——每条子 feature 进 cs-feat-design 时,roadmap 第 2 块的接口契约就是它的硬约束输入(不能违反,要改先回 roadmap update)。

为什么 roadmap 承载架构方案不放进 architecture/:cs-arch 守"只记现状不记计划"。前瞻性架构方案属于"还没落地、可能还会变"的事前规划,放进 architecture 会污染那份系统地图。等子 feature 真正落地,对应接口由 cs-feat-accept 提炼回 architecture/——roadmap 完成过渡使命后归档。

为什么单独一层:requirements 记"要什么"(愿景)、architecture 记"怎么搭"(结构)、roadmap 记"怎么分步实现"(执行)。把执行规划塞进愿景或结构文档会把"要什么"和"打算怎么实现"混起来——查不到系统真实能力,计划改一下又得改两份文档。

为什么文件夹不是单文件:拆解过程会产生草稿 / 调研 / 方案对比 / 白板转述,塞一份 md 会乱又舍不得删。每个 roadmap 一个子目录,主文档对外口径,旁边 drafts/ 随便堆。

共享路径与命名约定看 .codestable/reference/shared-conventions.md。主文档和 items 完整模板看同目录 reference.md。

---

适用场景

  • 用户描述"一眼看出做不完"的大需求("加权限系统"、"做通知中心"、"接 SSO")
  • cs-brainstorm 判为 case 3 移交过来(brainstorm 只做分诊,不做拆解)
  • 已有 roadmap 加新子 feature / 改依赖 / 调顺序 / 标废弃
  • feature-design 发现要做的事实际是多个 feature 集合,先退回拆

不适用:单 feature 能装下 → cs-feat;描述能力"是什么、边界" → cs-req;描述系统"结构怎么搭" → cs-arch;拍板长期规约 / 选型 → cs-decide。

---

模式分流

用户说什么模式
"拆一下 X 需求"、"开一份 X 的 roadmap"、"我想要一个 X 系统"new
"往 {已有 roadmap} 加子 feature"、"重排顺序"、"标 drop"update

判断不出问用户。

---

单目标规则

每次只动一份 roadmap。一次扔出"我想要 X 和 Y"先选一个,另一个下次。理由同 req / arch——一次吐多份用户 review 不过来。

---

目录结构

.codestable/roadmap/{slug}/
├── {slug}-roadmap.md       主文档:背景 / 范围 / 模块拆分(概设)/ 接口契约(架构层详设)/ 子 feature 清单 / 排期
├── {slug}-items.yaml       机器可读清单(feature-design 读、feature-acceptance 回写)
└── drafts/                 可选,调研 / 讨论 / 草稿

{slug} 小写字母 / 数字 / 连字符,和大需求一致(permission-system、notification-center)。平铺不嵌套 epic / sub-epic。drafts/ 按需建,AI 不强制归档。

---

工作流

Phase 1:锁定目标

模式 + 目标 + 范围。new 模式先敲定一个英文 slug(参考现有 req / arch slug 习惯)。

Phase 2:读取材料

共同必读:.codestable/attention.md + 用户素材 + roadmap/ 其他 roadmap(防重复)+ requirements/ 相关 req + architecture/ 相关 doc。

按情况读:

  • 相关 compound 沉淀:python .codestable/tools/search-yaml.py --dir .codestable/compound --query "{大需求关键词}"
  • 已有相关 feature 方案

update 额外:当前主文档全文 + items.yaml 当前状态 + 已启动 / 完成的子 feature 的 design / acceptance。

Phase 3:拆解与起草

按 reference.md "主文档结构"和"items.yaml 格式"写完整初稿不分批。

拆解纪律:

  1. 先做架构方案再拆 feature——顺序:先想模块拆分(概设第 3 节)→ 模块间接口 / 数据结构 / 协议(详设第 4 节)→ 才把方案分解成子 feature(第 5 节)。架构方案不清楚就硬拆 feature,结果是每个 feature 各自重新发明轮子、接口对不齐
  2. 接口契约要写到 feature 可拿来当硬约束的程度——函数签名 / 数据结构 / 协议字段 / 错误码这一级。讲不到这级回去想。无跨模块接口(如纯前端样式调整)就明确写"无跨模块接口"
  3. 每条子 feature 能当独立 feature 流程跑完——能单独写 design / 实现 / 验收。跑不下来颗粒度不对
  4. 依赖图必须是 DAG——A 依赖 B 写清楚,别循环
  5. 依赖关系要有具体理由——"B 依赖 A,因为 A 提供 XX 表结构"而不是"A 先做"
  6. 先列一条最小闭环——做完后能端到端跑通最窄路径的标第一条
  7. 明确不做的边界——用户脑子里的"权限系统"可能包括审计日志 / 数据脱敏,不打算覆盖就写进"明确不做"
  8. 不替用户决定优先级——技术依赖之外的排序让用户拍板

Phase 4:自查清单

review 前自跑一遍汇报处理:

  1. 模块拆分讲清了吗?每个模块职责一句话能说出来?
  2. 接口契约写到可执行程度了吗?feature-design 看完不需要回来问就能直接照着实现?
  3. 每条子 feature 的 slug 规范?(grep .codestable/features/ 确认不冲突)
  4. 每条描述一句话讲清楚?讲不清就拆得不够或 scope 太模糊
  5. 依赖关系是 DAG?有没有自指 / A→B→A 回环
  6. 最小闭环真的最小?第一条做完能独立给用户演示点什么?
  7. "明确不做"有没有写?没写就说"没有明确不做"
  8. 和已有 req / arch 有没有矛盾?有就写"和 req-X 冲突待用户决定",不偷偷选边
  9. update 专项:本次新加 / 改条目都有素材依据?凭空"加一条让看起来更完整"是漂移
  10. update 专项:改了接口契约的话,已 in-progress / done 的子 feature 受影响吗?影响到的列"观察项"提示用户

Phase 5:用户 review

主文档 + items.yaml 完整贴给用户。改到用户明确"可以了"。

Phase 6:落盘

new:建 .codestable/roadmap/{slug}/;写主文档(status: active / created / last_reviewed 当天);写 items.yaml(每条 status: planned、feature: null);validate-yaml.py 校验。

update:改主文档(last_reviewed 当天,结构性改动文末加变更日志);改 items.yaml 对应条目(drop 不删,status: dropped 留存理由);重新校验 yaml。

不改 requirements / architecture——roadmap 是规划层,那两层只描述现状。拆解过程发现 req / 架构过时,在主文档"观察项"记一句给用户,不顺手改。

---

和 feature 流程的衔接

feature 从 roadmap 起头

用户说"开始做 roadmap 里的 {子 feature}"时:

  1. cs-feat-design(或 ff / brainstorm)起 feature 目录
  2. design frontmatter 带 roadmap: {slug} + roadmap_item: {子 slug}
  3. items.yaml 对应条目 status: in-progress、feature: YYYY-MM-DD-{slug}

职责在 cs-feat-design 不在本技能。

feature-design 必须把 roadmap 接口契约当硬约束

roadmap 主文档第 4 节"接口契约"是 feature 的硬约束输入——不是参考,是不能违反、要改先回 roadmap update。这就是为什么 roadmap 要在拆 feature 前先把架构方案定下来:让多条 feature 并行 / 串行实现时对外接口对齐。

feature-design 发现接口契约不合理 / 漏了 / 描述不准 → 回 cs-roadmap update 改了再继续,不要在 feature 里偷偷绕开(绕开会让下一条同模块 feature 接到老契约导致二次冲突)。

acceptance 自动回写

cs-feat-accept 收尾时如果 design frontmatter 有 roadmap 字段就改对应 roadmap_item 的 status: done,同时同步主文档子 feature 清单勾选状态。职责在 cs-feat-accept 不在本技能。

roadmap 自身生命周期

  • 所有 items done / dropped 后主文档 status: completed,目录留作历史档案
  • 长期无进展:status: paused,主文档加理由

---

硬性边界

  1. 不写单 feature 内部实现细节——roadmap 写到"模块边界 / 接口契约 / 共享协议"为止,单模块内部怎么实现归 feature-design。判据:会被多个 feature 共同遵守归 roadmap,只在某个 feature 内部用归 feature-design
  2. 不改愿景和结构档案——不顺手改 requirements / architecture / 代码 / 已有 feature。问题记"观察项"
  3. 不替用户拍产品优先级——技术依赖外的排序让用户决定
  4. 单目标——一次只动一份
  5. 不发散——用户范围外问题记观察项不扩大
  6. 接口契约要么可执行级要么明确"无跨模块接口"——不允许"待定 / 后面再说"。含糊掉的接口在 feature-design 各条会各自补出来必然不一致

---

退出条件

  • [ ] 已锁定单一模式 + 单一目标
  • [ ] 主文档 frontmatter 完整(doc_type: roadmap / slug / status / created / last_reviewed / tags)
  • [ ] 主文档含:背景 / 范围与明确不做 / 模块拆分 / 接口契约 / 子 feature 清单 / 排期 / 观察项
  • [ ] 模块拆分节每个模块职责一句话讲清
  • [ ] 接口契约节写到 feature-design 可拿来当硬约束的级别(函数签名 / 数据结构 / 协议字段 / 错误码)或明确"无跨模块接口"
  • [ ] items.yaml 每条有 slug / description / depends_on / status / feature
  • [ ] 依赖图是 DAG 无循环
  • [ ] 最小闭环条目已标
  • [ ] items.yaml 通过 validate-yaml.py 校验
  • [ ] Phase 4 自查清单逐条跑过并汇报
  • [ ] 用户 review 通过
  • [ ] 没有顺手改 req / arch / 代码 / 已有 feature

---

和其他工作流的关系

方向关系
cs-req 配合req 记"为什么有这个能力"、roadmap 记"打算怎么分步做出来"。大需求下可能多份 req;缺 req 提示用户先 cs-req
cs-arch 配合architecture 记现状、roadmap 记若干步。读 arch 理解现状但不改它
cs-feat 下游每条子 feature 是未来一次 feature 流程的种子;起头时 design frontmatter 带 roadmap / roadmap_item
cs-feat-accept 回写方acceptance 自动改 items.yaml 为 done,本技能只定义格式不负责回写
cs-onboard 创建者建 roadmap/ 空目录
cs-brainstorm 上游case 3 移交本技能,带"真问题 / 大致范围 / 可能子模块"一句话汇总。本技能不重复分诊直接拆

---

常见错误

  • 跳过架构方案直接拆任务——上来就列子 feature,模块边界 / 接口没想,feature-design 各自发明轮子
  • 接口契约写得含糊——"两边商量"、"待定"、"用一个统一的事件总线"——讲不到字段 / 签名 / 协议级。feature-design 接到这种没法当硬约束
  • 把单 feature 内部细节写进 roadmap(某模块内部怎么分文件 / 用哪个库)——归 feature-design
  • 颗粒度失衡——一条装得下三个独立功能、另一条只改个配置
  • 依赖关系靠脑补——讲不清为什么依赖
  • 替用户排优先级
  • 和已有 req / arch 冲突不停下——自己选一边掩盖真实分歧
  • 一次做多份 roadmap
  • 顺手改 req / arch
  • drop 条目直接删——历史丢失
  • roadmap 跑偏成给单条子 feature 写详细方案
  • update 改接口契约不评估存量影响——已 in-progress / done 的 feature 没人看到契约变了

Score

0–100
63/ 100

Grade

C

Popularity15/30

998 installs — growing adoption.

Completeness27/30

Documented: full SKILL.md body, description, one-line install. Missing: category/license metadata.

Trust15/25

Community skill with a public GitHub source repository you can review.

Freshness6/15

No update timestamp is tracked for this skill in our catalog.

Scored automatically from popularity, completeness, trust, and freshness — computed only from data in our catalog, never fabricated.

Proud of your score? Add this badge to your README.

Paste a snippet into your GitHub README. The badge updates automatically and links back to this page.

Cs Roadmap skill score badge previewScore badge

Markdown

[![Cs Roadmap skill](https://www.remoteopenclaw.com/skills/liuzhengdongfortest/codestable/cs-roadmap/badges/score.svg)](https://www.remoteopenclaw.com/skills/liuzhengdongfortest/codestable/cs-roadmap)

HTML

<a href="https://www.remoteopenclaw.com/skills/liuzhengdongfortest/codestable/cs-roadmap"><img src="https://www.remoteopenclaw.com/skills/liuzhengdongfortest/codestable/cs-roadmap/badges/score.svg" alt="Cs Roadmap skill"/></a>

Cs Roadmap FAQ

How do I install the Cs Roadmap skill?

Run “npx skills add https://github.com/liuzhengdongfortest/codestable --skill cs-roadmap” in your terminal. The skill is added to your agent's skills directory and picked up automatically on the next run — no restart or extra configuration needed.

What does the Cs Roadmap skill do?

把"大到塞不进单个 feature"的需求做成完整事前规划:概设 + 接口契约 + 子 feature 拆解清单,放在 `.codestable/roadmap/{slug}/`。两种模式 new / update。触发:用户说"我想要一个 X 系统"、"帮我把这块需求拆一下"、"开一份 roadmap",或 feature-design 阶段发现需求太大。 The full SKILL.md on this page shows the exact instructions the skill gives your agent.

Is the Cs Roadmap skill free?

Yes. Cs Roadmap is a free, open-source skill published from liuzhengdongfortest/codestable. As with any third-party skill, review the source repository before installing it into an agent with sensitive access.

Does Cs Roadmap work with Claude Code and OpenClaw?

Yes. Skills use the portable SKILL.md format, so Cs Roadmap works with Claude Code, OpenClaw, Codex, Hermes, and any other agent that reads SKILL.md skills.

Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →
Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →
Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger →
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger →
Turn any website into LLM-ready data with Firecrawl logoTurn any website into LLM-ready data with Firecrawl

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits plus 10% off through our link.

Try Firecrawl free →
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now →
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free →

Categories

Command ExecutionPrompt Injection
View on GitHub

Recommended skills

Browse all →
find-skills logo

find-skills

vercel-labs/skills

2.7M installsInstall
frontend-design logo

frontend-design

anthropics/skills

720K installsInstall
grill-me logo

grill-me

mattpocock/skills

701K installsInstall
agent-browser logo

agent-browser

vercel-labs/agent-browser

596K installsInstall
grill-with-docs logo

grill-with-docs

mattpocock/skills

594K installsInstall
vercel-react-best-practices logo

vercel-react-best-practices

vercel-labs/agent-skills

591K installsInstall

Browse

Skills by category

Frontend250Git198Data154Testing120Design105Docs103Security96Automation87Backend76Devops37Productivity29Mcp23

Related guides

Hand-picked reading to help you choose, install, and use agent skills.

Guide10 Openclaw Skills Every Nextjs Developer NeedsGuideBest Openclaw Skills 2026GuideHow To Evaluate Openclaw Skill Before Installing

Remote OpenClaw

AI agent skills directory, marketplace, and workflow hub for OpenClaw, Hermes Agent, Claude Code, Codex, and MCP-powered operator stacks.

The Agent Stack: weekly agent tooling digest, free.

Explore

  • Home
  • Skills Directory
  • Claude Code Skills
  • Codex Skills
  • MCP Clients
  • Marketplace
  • Hermes Ecosystem
  • Free guide
  • Learn
  • OpenClaw for Creators
  • OpenClaw for Founders
  • Blog
  • The Agent Stack (Digest)

More

  • Submit a Tool
  • Advertise
  • Playbook
  • Free Tools
  • API
  • Shipping
  • Contact
  • Terms
  • Privacy

Know a company that should advertise here? Refer them and earn 10% — up to $300 per referral.

© 2026 Remote OpenClaw
Fazier badgeFeatured on Twelve ToolsFeatured on Wired BusinessRemote OpenClaw - Featured on AI Agents DirectoryListed on Turbo0Featured on Uneed