Gearcoleco Romhacking
drhelius/Gearcoleco
Hack, modify, and translate ColecoVision and Super Game Module ROMs using the Gearcoleco emulator MCP server.
Author or revise a cloud-api-operations recipe (config/source/skills/cloud-api-operations/references/recipes/).
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoring --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .claude/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .claude/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoringType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoring --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .agents/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .agents/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoring --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .cursor/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .cursor/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git --path skills/cloud-api-recipe-authoring--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoring --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .gemini/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .gemini/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoringInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .github/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .github/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install TencentCloudBase/CloudBase-AI-Toolkit cloud-api-recipe-authoring --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/TencentCloudBase/CloudBase-AI-Toolkit.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/cloud-api-recipe-authoring .opencode/skills/cloud-api-recipe-authoring && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "cloud-api-recipe-authoring" agent skill from https://github.com/TencentCloudBase/CloudBase-AI-Toolkit/tree/main/skills/cloud-api-recipe-authoring into .opencode/skills/cloud-api-recipe-authoring/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cloud-api-recipe-authoring", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
cloud-api-recipe-authoringAuthor or revise a cloud-api-operations recipe (config/source/skills/cloud-api-operations/references/recipes/).
Cloud API Recipe Authoring is an agent skill from TencentCloudBase/CloudBase-AI-Toolkit. Author or revise a cloud-api-operations recipe (config/source/skills/cloud-api-operations/references/recipes/). Encodes why the long tail of control-plane capabilities is deliberately NOT wrapped as MCP/CLI tools, the API-first admission test for whether a capability deserves a recipe, the negative and positive lists, the fixed recipe structure, the writing rules distilled from writing-for-agents, the user-facing copy red lines, and the product regeneration/verification pipeline. Also routes contributors (and…
Its SKILL.md is about 3.7k tokens, which your agent loads only when the skill is triggered. The skill folder holds 4 other files, including scripts and reference files (for example `references/contributing.md`).
It sits in Agent Workflows, covering Brand voice and tone and MCP servers. It works with Model Context Protocol. The repository describes itself as: Backend for AI coding agents on CloudBase — database, auth, functions via Plugin, Skills & MCP. The licence is MIT.
3 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit 21af91c. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships 1 file in scripts/ (JavaScript), which the agent can run.
Shell commands in SKILL.md call:
gitnpmnodeFrom the folder's file list and the shell code blocks in SKILL.md.
Hosts in commands or code, which the agent is likely to contact:
docs.cloudbase.nettcb.cloud.tencent.comFrom URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Cloud API Recipe Authoring loads about 3.7k tokens when it runs, and up to ~5.4k if it reads all its reference files. Until then it costs about 214 tokens; SKILL.md has 869 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
The full file from TencentCloudBase/CloudBase-AI-Toolkit at commit 21af91c, republished under its MIT licence (© TencentCloudBase). 869 words, ~3,722 tokens.
.claude/skills/cloud-api-recipe-authoring/SKILL.md (or your agent's skills folder). This skill also uses 2 other files; get the full folder from GitHub.目标:让**「AI 可以 100% 操作云开发」可验证 —— 每个高频管控面场景都有一篇能照着跑通的 recipe,且每篇都建立在公开 API** 之上。
config/source/skills/cloud-api-operations/references/recipes/README.md 的目录表与状态列./references/contributing.md 给他产品排期里「对外开放 API」通常不优先,于是有一个固定落差:控制台上了功能,AI 却操作不了。因果链是单向的 —— 没有公开 API,CLI / MCP 就无从封装,AI 只能用鼠标。
补齐落差有两条路,我们不走「全都封装」那条:
| 做法 | 代价 |
|---|---|
| 把能力都补成 API、再全部封装成工具 | 工具面持续膨胀:模型选错工具的概率上升,工具描述吃掉上下文预算,每个工具都要跟 API 变更一起维护 |
| 只封装核心,长尾走 recipe | 要多维护一份文档,但工具面稳定,长尾能力照样能被 AI 操作 |
核心能力已经封装过了(query* / manage* 工具家族 + CLI),剩下的大量接口不需要封装。一个能力该去哪:
| 能力形态 | 归属 |
|---|---|
| 有公开 API + 高频核心 | 已在 MCP / CLI 工具里 —— 直接用,不要另写 recipe |
| 有公开 API + 长尾(串联多 Action、有非平凡前置、有踩坑或回查义务) | 写 recipe(本 skill) |
| 有公开 API + 单接口、读文档即可调 | 既不封装也不写 recipe —— agent 现场查索引 |
| 没有公开 API(只在控制台可见) | 谁也封装不了 —— 产品缺口,推产品先开放 API;这里也没有 recipe 可写 |
最后一行是硬约束:没有公开 API,能力对 AI 的可用性就是 0,这不是「我们封装得不够」的问题。
三条必须同时满足:
https://docs.cloudbase.net/ai/cloudbase-ai-toolkit/api-reference.md 或产品官方文档)。不满足时的处理:
config/source/skills/(那是对外发布面)specs/cloud-api-skill/pending-recipes/(内部目录,不进版本控制),正文顶部写清「为什么未通过 + 解封条件」;外部贡献者不提 PR,按 ./references/contributing.md 开 Issue 说明踩过的实例见下文「反面 case」—— 驳回后不进公开集,也不要把归档内容抄回来。
SKILL.md §1 的发现流程现场查query* / manage* 家族)→ 路由到工具,不写文档DescribeDBInstances)集成中心 UserKey 密钥对管理 —— 全文在内部归档 specs/cloud-api-skill/pending-recipes/integration-userkey.md(不进版本控制,外部读者看不到内容,按下文「驳回后归档页写三样」自行落笔)。
它当时过了正向清单的两条:有非平凡前置(password 字段加密)、有实测踩坑(脱离 SDK 会明文落库)。但三条硬门槛一条都不过:
| 硬门槛 | 实际 |
|---|---|
| 有公开云 API | 6 个 Action(GetUserKeyList / CreateUserKey / …)在公开概览 0 命中 |
| 语言无关可调用 | 唯一通路是 @cloudbase/manager-node 的封装;Python / Go / Java 用户没有对应入口 |
| 写操作能独立完成 | password 的加密密钥由 SDK 内部从七彩石取(QueryRainbowConfig),公开面没有入口 → 裸调云 API 无法加密,写路径走不通 |
要记住的是漏判机制,不是这条结论:把「官方 SDK 文档里写全了、本地跑得通」当成了「能力在公开 API 契约内」。SDK 文档公开 ≠ Action 公开。
写前自检:动笔前拿目标 Action 名去公开概览 / 官方接口文档搜一次(入口见上文准入 §1)。0 命中就停手,不进公开集。
驳回后归档页写三样(照 pending-recipes/ 那份抄):① 为什么未通过,逐条对到上面三条门槛;② 解封条件——满足其一即可移回公开集;③ 显式禁止「复制回公开 skill」。
使用者看到的是「点一下授权链接」;我们要判断的是要不要让这一步消失。三种处置:
| 处置 | 什么情况 | 怎么做 |
|---|---|---|
| 预置进角色族 | 能力通用、高频、且在 TCB 产品语义内(TCR 拉镜像、CloudBase Run 访问 VPC/CVM 属这类) | 推产品给 TCB_QcsRole 挂预设策略,命名沿 QcloudAccessForTCBRoleIn<X> → 用户零操作 |
| 让用户按需追加 | 长尾能力、跨产品,或预设策略粒度太粗(QcloudMonitorFullAccess 是全读写) | 给一键授权链接,写进 recipe 的 前置权限。默认走这条 —— 把大权限挂在默认角色上,等于替所有用户扩大默认授权范围 |
| 都不做 | 能力没有公开 API(准入驳回那一类) | 到此为止,不写 recipe 也不预置;当产品缺口反馈 |
判断顺序:先过准入(没有公开 API 就停),再看值不值得进角色族;两条都不过就只留一键授权链接。
动笔预置前先做一次身份校验:确认这条通路的凭据真的以该角色发起。auth get_temp_credentials 导出的是 Web / device 登录的同一份临时密钥(不是另起一次角色假定),所以只有用「权限收窄到刚好没这个 Action」的身份实测报出 UnauthorizedOperation,才能判定角色策略在这条通路上生效。判不出来就给一键授权链接、不预置。
一个已观察到的线索:账号级 device 登录的凭据能过 cam:CreateRole、cdn:DescribeDomains,而同一账号 TCB_QcsRole 的挂载策略里都没有这两条 —— 倾向说明它取的不是该角色。但这个结果用主账号登录也会得到,必须换窄权子账号复测才算数;没复测前,对外只写「按凭据身份判断」,不要对外断言 CLI / MCP 走角色。
已落定的归属(按能力族,不按产品):
| 能力 | 处置 | 依据 |
|---|---|---|
| 云开发自己的域名(静态托管、网关自定义域名的绑定 / 解绑 / 查任务) | 预置——已在角色族 | 走 tcb:* 与 cdn:Tcb*,属 TCB 产品语义内 |
| 证书只读(列证书 / 详情 / 校验链) | 预置——已在角色族 | 绑自定义域名要在证书列表里挑一张,是必要读权限 |
| 云监控告警族 | 预置——已在角色族 | 随基座策略下发 monitor:*,不单独挂 QcloudMonitorFullAccess |
通用 CDN 加速域名(AddCdnDomain 等) | 让用户按需追加 | 跨产品长尾;QcloudCDNFullAccess 粒度太粗 |
| DNS 解析(DNSPod) | 让用户按需追加 | 跨产品长尾:解析记录操作与云开发无产品语义绑定 |
| 证书写(申请 / 续期 / 删除 / 部署) | 让用户按需追加 | 同上;QcloudSSLFullAccess 是全读写,粒度偏粗 |
角色挂载策略会随产品迭代变动,上面这张表可能过期 —— 以 ListAttachedRolePolicies + GetPolicy 读到的实际策略内容为准,不要拿历史结论当事实。
recipe 里出现的 service 要能在 callCloudApi 直接调通。白名单是 mcp/src/tools/capi.ts 的 SERVICE_VERSIONS(枚举 + 官方版本映射)。写 recipe 时若发现目标产品不在表里,先补表再写 recipe:单版本产品写 service: ["YYYY-MM-DD"],多版本产品把官方在用版本都列进数组(缺省时调用方必须显式传 version)。版本号取自官方 SDK 目录 tencentcloud/<service>/v<YYYYMMDD>,不要凭记忆填。
改完 SERVICE_VERSIONS 要同步两处再跑产物链:使用者侧的完整索引 config/source/skills/cloud-api-operations/references/service-versions.md(按产品线分组的分组表),以及 references/calling-methods.md §1 的「最常用的几个」小表(只在核心产品变动时动这张)。
链接拼法、角色载体读法、策略覆盖范围见对外 skill 的 config/source/skills/cloud-api-operations/references/calling-methods.md §3 —— 维护者视角的内容只写在本 skill(对外 skill 只写「怎么点一下授权」,不写这一节)。
准入过了不等于能写 —— 先拿真账号把只读路径跑通,再决定哪些进主序列、哪些进「踩坑清单」。
SERVICE_VERSIONS 时先补表(mcp/src/tools/capi.ts),再重建本地产物:跑 build:webpack,不要跑 npm run build —— 它的 prebuild = rm -rf dist 会被 safe-delete 钩子拦下,整条链断在第一步。版本号取自官方接口文档页或 SDK 目录 tencentcloud/<service>/v<YYYYMMDD>,两处对得上才填。callCloudApi 有环境绑定门禁:账号级登录后首个调用仍会返回 ENV_REQUIRED(「已登录,但尚未绑定环境」)。先 auth(set_env, envId=…) 绑一个正常环境即可 —— 备案这类账号级业务与环境无关,绑定只是为了过门禁。mcp/dist/cli.cjs(initialize → tools/call),比挂进客户端快得多;本地 MCP 起法与探针脚本见 skill cloudbase-mcp-local-probe。Describe* / Check* / Validate*,写操作一律留给使用者手动执行。写路径如果必须靠订单流程内的参数才能发起,正好说明它不进主序列。🟡 部分实证。挑按量计费、可重建、不影响他人的环境,走完整条序列(提交 → 轮询进度 → 回查结果 → 改回原值并再次确认);两个方向都要跑,升与降常走不同分支、耗时也不同(实测升 174 s / 降 67 s,同规格同地域)。真跑还会暴露只读永远看不到的东西:任务步骤表会边跑边长、入参与任务详情对同一个字段用两套枚举。DryRun / 预检),可以按只读口径实测 —— 但两个条件都要满足:(1) 官方文档或控制台自己的实现明确说明它不产生订单、不发起变更;(2) 参数挑「目标值 = 当前值」,万一开关没生效,也只会命中幂等分支、不会真改。跑完必须再读一次现状确认没变(例:变配预检后重读实例属性,比对规格与 UpdateTime)。src/framework/routes/groups/** 是路由真源(拿到页面 hash 路径),src/services/capi-action-definitions/** 是「控制台调哪个云 API、每个字段什么语义、返回体的空值约定」的真源;页面自己的 hook / utils 还能看到数据来源与单位换算(字段名相同但单位不同,多半只有这里写着)。写 recipe 前先读这三处,比只读官方文档多拿到「真实调用序列 + 字段语义 + 踩坑」三样东西。queryCloudRun / manageCloudRun 这类工具的 action 列表就是它的能力边界(列表里没有的才轮到 callCloudApi)。一个能力同时有专用工具和云 API 时,recipe 里只写专用工具;只有专用工具没接的那一段才写裸云 API,并在文中说清为什么绕开(例:云托管固定 IP 在 tcbr 有 Action,但两个云托管工具的 action 列表里都没有)。tcbr)的 EnvId,但云托管要单独开通。写「前置权限」表时按能力分 service,正文里点明「这个能力可能还没开通、怎么判、怎么开」,否则读者会把空值当成「已开通且未开启」。Status 都是空串)。拿它当判据前,先找到能把状态区分开的那个调用,把它排进序列的前面,否则读者拿到空值也不知道下一步该干什么。判据字段有取值域时按取值域写全,别用「空值 = 关闭」这种近似:类型定义里带枚举注释的字段(例 status: string; // normal | isolate)就写成「normal = 已开启,其余值含空串 = 未开启」,否则读到第三个值的人不知道该怎么判。tcb/DescribeICPResourcesInfo 返回 success + 空数组(像「够格」)、tcbr/DescribeStableIP 返回 success + 空 Status、queryCloudRun(action="envStatus") 返回 unopened,同一批探测里只有 tcb/DescribeEnvInfo 报 env not found。动手验证前先用一个「一定会报错」的接口探目标可达性,否则整轮实测的结论都建在空值上。ba 这类操作级授权产品,账号级凭据确实可能整族被拒(九个只读接口全部 UnauthorizedOperation,action 名被正常识别、卡在策略层)。但同一批接口在业务主体所属账号的凭据下是正常返回数据的 —— 两边都实测过,所以真正的变量是凭据与目标是否同账号。「前置权限」里既不要写「账号读自己的数据一般不需要额外策略」这种没验证过的乐观断言,也不要把一次被拒写成「该产品要额外策略」;写清账号这个变量,并给出换对账号后的预期。验证前先确认凭据与目标同账号,否则整轮结论都是「账号不对」造成的假象。missing the required parameter Skey / IcpOrderID / Uin 这类只能从内部流程拿到的参数,就等于宣告该接口无法独立使用 —— 它该写进「踩坑清单」的反面做法,而不是主序列。https://tcb.cloud.tencent.com/dev?envId=<envId>#/<path> 拼。控制台路径的唯一真源是 config/source/skills/cloudbase-platform/SKILL.md 的 Entry points 清单 —— 那里没有的路径先补那一份,再在 recipe 里引用,否则两处各写一份必然漂移。同一条判据往往有「看起来该在这儿」的错页(例:备案准入的固定 IP 开关在备案管理页,环境设置页没有卡片),把这类对照写进「踩坑清单」,比只给一条正确链接更省使用者的时间。顺序固定:When to use → 前置权限 → 接口序列 → 踩坑清单 → 验证步骤(+ 可选 入口)。对外发布的是这套顺序本身,别增删标题。
| 节 | 写什么 | 完成条件(可判定) |
|---|---|---|
When to use | 出现什么症状就该翻这篇(可判定的触发条件,不是主题概括);末尾一句划界:本篇不管什么、该转哪去 | 触发条件能对上用户手里的报错/现象 |
前置权限 | 需要的 service / version、凭据身份(账号级 / 环境级 / 服务角色)、需要哪几个 CAM 策略;能直接用的官方文档入口 | 权限不足时能直接给出一条可点击的一键授权链接(角色名 + 策略名 + principal),而不是"去 CAM 控制台追加策略" |
接口序列 | 最小可跑序列(代码块或步骤表),每步带返回要点;方法/参数只在一处写全 | 每步的返回值都能对应到下一步的入参;表格每一步都带 service 列(表头写 service 或 service / version),用到多个 service 时分节写、每节表头都要有该列 —— 漏掉这列,读者只能靠猜这一步是谁的 API |
踩坑清单 | 表格:坑 / 现象(贴真实报错原文)/ 正确做法 | 每行都能对上一条真实报错或可复现现象 |
验证步骤 | 调用后怎么自查(只读自检、写后回查、幂等查重) | 每条都有明确通过判据(字段有值 / 计数一致),不用「合理」这类词 |
| 杠杆 | 本仓库的落地 | 反例 |
|---|---|---|
| 单一可信源 | 序列只在代码块或表里写一次;索引表不重复它的内容 | 同一序列「表里一遍、代码里一遍」 |
| context pointer | 标题与首句写症状,让 agent 能判断「要不要翻这篇」 | 标题写成主题词(「关于告警」) |
| 信息层级 + progressive disclosure | in-file step > in-file reference > disclosed reference:只有某一支才需要的细节,下沉到分支里或指向可查处,不要塞进主序列 | 把 AES/IV/base64 全套写进正文,而只有「脱离 SDK 自己实现加密」才用得到 |
| co-location | 相关的东西放同一节,少造标题 | 「因为这篇要解决什么」+「能力入口」两个标题讲同一件事 |
| 完成条件 | 每节给可判定判据;要求 legwork 就明说读什么 | 「TotalCount 合理」无法判定,会诱发提前收工 |
| leading word | 复用现成紧凑词,别自造 | 把「SDK 取不到密钥时静默明文写入」起名 明文落库,与「密文落库自检」构成同一词汇对 |
| 正面表述 | 写「改查 X」「走 Y」,别写一串「不要…」(禁用词会把行为拖进上下文) | 「不要猜 Action 名、不要用索引、不要…」 |
| 环境即真源 | 命令与路径写成能在当前环境直接执行的;能本地查到的不要缓存 | 缓存 node_modules 里就能读到的实现细节 |
| 剪枝 | 删 no-op、删沉渣、删与别处重复的内容 | 同一事实在 SKILL.md / README / recipe 三处各写一遍 |
recipes/README.md 目录表的「状态」列(✅ 生产实证(日期) / 🟡 部分实证(日期) / 🟡 文档核对,未实跑),SKILL.md 条目与 recipe 正文都不重复✅:一篇里「跑过」与「没跑过」混着是常态 —— 写操作要真实破坏、权限或计费限制都可能挡住。✅ 要求整条接口序列端到端跑过;只跑到一部分(只跑通其中一条路径、或只跑了只读步骤)就用 🟡 部分实证,并把没跑到的部分写在状态列里,使用者才知道该对哪几条保持谨慎。整篇贴 ✅ 的代价与给整节贴「(全部实测)」相同:后来补进的未验证内容会继承它没挣到的置信度。(2026-09-24 教训:两篇 recipe 都只有部分路径真跑过,先误贴 ✅,自查时才发现 —— 写完状态列要回头拿「这条序列里哪几步真的执行过」核一遍。)✅ 一起被高置信采信。(2026-09-25:往 PG recipe 里补「无需升级套餐」时带进了控制台的资源点单价常量,正文标明是界面口径、状态列另点一条未实跑。)config/source/skills/cloud-api-operations/references/recipes/README.md,改结构/红线时两处一起看recipes/README.md 目录表:序号 + 场景名 + 文档链接 + 状态要加一篇 recipe 的完整流程、PR 该写什么、以及提 PR 前必须做的三层自测(机器校验 / 端到端实跑 / 多模型冷启动)见 ./references/contributing.md。
从 raw 链接直接读本文件时:相对路径接在 https://raw.githubusercontent.com/TencentCloudBase/CloudBase-AI-Toolkit/main/ 之后即可取到对应文件。
export PATH=/usr/local/bin:$PATH # 缺它 → git-lfs 钩子报错,&& 链会静默断掉
NODE=/Users/bookerzhao/.workbuddy/binaries/node/versions/22.22.2-3/bin/node
$NODE scripts/generate-prompts-data.mjs && $NODE scripts/generate-prompts.mjs
$NODE scripts/sync-claude-skills-mirror.mjs && $NODE scripts/build-compat-config.mjs
$NODE scripts/check-prompts-sync.mjs && $NODE scripts/sync-claude-skills-mirror.mjs --check
$NODE scripts/diff-compat-config.mjs # Has blocking diff: NO
$NODE skills/cloud-api-recipe-authoring/scripts/check-recipes.mjs # 表格结构 + 编号交叉引用动过接口序列的结构(删节、加节、重排步骤号)之后,必跑 check-recipes.mjs。 步骤号在两个互不相干的位置各写一遍 ——「步」列的单元格、正文里的「步骤 N」句子 —— 改一处不会让另一处报错,只会让读者照着一个不存在(或指向别的步骤)的编号去找。脚本还会检查表格列数(含 \| 转义、分隔行、表格后空行)与反引号配对,退出码非 0 即有问题。不要用眼睛代替它:编号差一位看起来完全正常。
新增或删除文本面都会让 compat-diff 报 blocking,必须刷 config/source/editor-config/compat-baseline.json
新写一篇 recipe = 新增文件,只能用全量刷新。 update-compat-baseline.mjs --only <skill> 只改已存在的 key;匹配到新文件时会把它记进 unseen 并提示走全量 —— 不会替你加进去(existence 级变更只能整体重算)。所以:
--only cloud-api-operations),也避免顺手洗白别人的漂移npm run update:compat-baseline全量刷新后先看 diff 范围再提交,别默认它吞了别人的漂移。实测(2026-09-24,新增一篇 recipe 后全量刷新):baseline diff 的改动路径全部落在本次动过的两个 skill 上,无一条落在别的 skill —— 因为分叉点上的 baseline 与源是一致的。判据是可机械检查的:把 diff 里的路径去重,只应出现你改过的 skill 名;出现第三个就停下查那个 skill 的源与 baseline 谁过期。
新 recipe 还要带上两份派生产物:config/.claude/skills/** 镜像与 doc/prompts/cloud-api-operations.mdx。跑测试时注意 /usr/local/bin/node 会让 vitest 直接起不来,绕法见 skill cloudbase-mcp-artifact-pipeline
校验完成前不要宣称改完:三项预检全过 + git diff 里非目标改动行为空
grep:本会话实测会静默返回空且 exit 0(假阴性),用内置检索工具/usr/local/bin(git-lfs 在那里),git checkout <file> 会返回非零并断掉 && 链git log -S"<原句>" 找引入提交,再翻当日的实测记录)。实测条件会变(账号不同、区域不同、构建版本不同),把「条件不同导致复现不了」写成「标注无证据」,等于自己造一次回退。本轮就发生过:一条值级「实测」标注差点被当成无证据撤掉,实际在当轮探针记录里能查到真实返回。config/.claude/ 镜像,还要清掉 SKILL.md 与 README 里的引用,再跑产物链scripts/check-recipes.mjs 检出,别靠复查时"看着没问题"© TencentCloudBase, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 2 other files (scripts, references) in skills/cloud-api-recipe-authoring of TencentCloudBase/CloudBase-AI-Toolkit.
Open the folder on GitHubat commit 21af91c
Cloud API Recipe Authoring next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Cloud API Recipe Authoring this skillTencentCloudBase/CloudBase-AI-Toolkit | 1.1k | — | ~3.7k | Automated safety check: Pass | MIT | |
| Gearcoleco Romhackingdrhelius/Gearcoleco | 141 | — | ~3.9k | Automated safety check: Pass | GPL-3.0 | |
| Scenario Orbit Viewsscenario-labs/skills | 913 | 1 repos | ~2.3k | Automated safety check: Pass | MIT | |
| Nemoclaw User GuideNVIDIA/skills | 3.5k | — | ~1.2k | Automated safety check: Pass | Apache-2.0 | |
| MCP Server Builderanthropics/skills | 180k | 64 repos | ~2.3k | Automated safety check: Pass | Apache-2.0 | |
| MCP Server BuildershareAI-lab/learn-claude-code | 78k | 5 repos | ~1.2k | Automated safety check: Pass | MIT |
drhelius/Gearcoleco
Hack, modify, and translate ColecoVision and Super Game Module ROMs using the Gearcoleco emulator MCP server.
scenario-labs/skills
A skill your agent uses when a single image of a character, creature, prop, vehicle, or building must be seen from new camera angles through the Scenario MCP server: turnarounds, a 360 orbit, views…
NVIDIA/skills
Guides human users' AI agents to the NemoClaw docs MCP server and canonical Fern documentation in Markdown form.
anthropics/skills
Guides the design and implementation of Model Context Protocol servers in TypeScript or Python, from tool naming and error messages to evaluation.
shareAI-lab/learn-claude-code
Walks through building MCP servers in Python or TypeScript that expose tools, resources and prompts to Claude, with templates, registration and testing.
anthropics/claude-plugins-official
Explains how to bundle Model Context Protocol servers in a Claude Code plugin, covering config files, stdio, SSE, HTTP and WebSocket server types, and authentication.
TencentCloudBase/CloudBase-AI-Toolkit
A skill your agent uses for Node.js backend AI via @cloudbase/node-sdk (=3.16.0) — cloud functions, CloudRun, Express/Koa/NestJS, serverless APIs, scheduled jobs, LLM proxies, agent orchestration.
TencentCloudBase/CloudBase-AI-Toolkit
CloudBase official HTTP API client guide. An agent skill from TencentCloudBase/CloudBase-AI-Toolkit.
TencentCloudBase/CloudBase-AI-Toolkit
Analyze, standardize, validate, and sync locally maintained skills into agent skill directories with a skills CLI-aligned workflow.
TencentCloudBase/CloudBase-AI-Toolkit
Build production-ready AI agent backends using the CloudBase Agent Python SDK — create agents with LangGraph/CrewAI/LlamaIndex, serve them via FastAPI with AG-UI protocol streaming +…
TencentCloudBase/CloudBase-AI-Toolkit
A skill your agent uses when you develop, design, build, deploy, debug, migrate, or troubleshoot CloudBase (腾讯云开发, 云开发, TCB, 微信云开发) projects — Web, 微信小程序, 小程序, uni-app, mobile (iOS, Android…
TencentCloudBase/CloudBase-AI-Toolkit
CloudBase declarative deployment from a cloudbaserc config (声明式部署, 配置式部署, cloudbaserc 部署) through the deployBuild / deployPlan / deployApply MCP tools.
Works with
Categories
Author or revise a cloud-api-operations recipe (config/source/skills/cloud-api-operations/references/recipes/). Cloud API Recipe Authoring is an agent skill from TencentCloudBase/CloudBase-AI-Toolkit. Author or revise a cloud-api-operations recipe (config/source/skills/cloud-api-operations/references/recipes/).
Cloud API Recipe Authoring fits situations like: adding a recipe; judging whether a scenario belongs in the recipe set; updating the recipes index; re-checking existing recipes against the admission test.
Run `npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a claude-code`. Or copy the skill folder (skills/cloud-api-recipe-authoring in TencentCloudBase/CloudBase-AI-Toolkit) into .claude/skills/cloud-api-recipe-authoring in your project. Claude Code loads it when a task matches its description.
Run `npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a codex`. Or copy the skill folder (skills/cloud-api-recipe-authoring in TencentCloudBase/CloudBase-AI-Toolkit) into .agents/skills/cloud-api-recipe-authoring in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add TencentCloudBase/CloudBase-AI-Toolkit --skill cloud-api-recipe-authoring -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cloud-api-recipe-authoring, .gemini/skills/cloud-api-recipe-authoring, .github/skills/cloud-api-recipe-authoring and .opencode/skills/cloud-api-recipe-authoring in your project.
Going by SKILL.md and its folder, Cloud API Recipe Authoring needs JavaScript for the scripts in its folder and the command-line tools its instructions call (git, npm and node). Our summary lists: Python 3; Node.js.
SKILL.md names 2 domains. In commands or code: docs.cloudbase.net and tcb.cloud.tencent.com; the agent is likely to contact these when it follows the instructions. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
Cloud API Recipe Authoring is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 3.7k tokens (SKILL.md is roughly 15k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 1.6k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Cloud API Recipe Authoring: Gearcoleco Romhacking (drhelius/Gearcoleco, 141 stars), Scenario Orbit Views (scenario-labs/skills, 913 stars), Nemoclaw User Guide (NVIDIA/skills, 3.5k stars) and MCP Server Builder (anthropics/skills, 180k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
TencentCloudBase (a GitHub organization) maintains it in TencentCloudBase/CloudBase-AI-Toolkit, which has 1,133 GitHub stars. The repository holds 49 skills in this directory. The repository was last updated on October 7, 2026.
Source: TencentCloudBase/CloudBase-AI-Toolkit on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.