A ClawdHub skill for managing recipes and grocery lists via cooking-recipe.
Data routing + security disclosure (mandatory)
- - Shared backend/API target: INLINECODE0
- Auth and recipe/grocery API traffic goes to configured
CONVEX_URL. - Shared backend usage is blocked unless
ALLOW_DEFAULT_BACKEND=true. - Never claim local-only processing when backend calls are involved.
Use when / don't use when (routing guardrails)
Use when:
- - User asks to add/list/search/show/delete cooking-recipe recipes.
- User asks to create/view/update cooking-recipe grocery lists.
- User asks to authenticate for cooking-recipe commands.
Do NOT use when:
- - User asks for general meal ideas (no cooking-recipe action requested).
- User asks for health/nutrition analysis outside stored cooking-recipe data.
- User asks to manage Apple Reminders/Notes/other systems.
Setup
- 1. Install:
CODEBLOCK0
- 2. Configure environment:
CODEBLOCK1
- 3. Environment variables:
-
CONVEX_URL (
required, recommended self-hosted backend)
-
ALLOW_DEFAULT_BACKEND=true (only if intentionally using shared backend)
- OAuth provider creds:
- GitHub (required):
AUTH_GITHUB_ID,
AUTH_GITHUB_SECRET,
GITHUB_CALLBACK_URL
- Google (optional):
AUTH_GOOGLE_ID,
AUTH_GOOGLE_SECRET,
GOOGLE_CALLBACK_URL
- Apple (optional):
AUTH_APPLE_ID,
AUTH_APPLE_SECRET,
APPLE_CALLBACK_URL
-
CLAWDBOT_URL (required, bot callback host)
Commands
- - INLINECODE15
- INLINECODE16
- INLINECODE17
- INLINECODE18
- INLINECODE19
- INLINECODE20
- INLINECODE21
- INLINECODE22
- INLINECODE23
- INLINECODE24
- INLINECODE25
- INLINECODE26
Output templates
Use response formatting templates from references/output-templates.md.
Security checklist before replying
- - If command failed due to auth/config, provide exact next step.
- If backend is relevant, keep disclosure truthful (shared default vs override).
- Never expose secrets/tokens in output.
技能名称:cooking-recipe
详细描述:
一个通过cooking-recipe管理食谱和购物清单的ClawdHub技能。
数据路由 + 安全披露(必填)
- - 共享后端/API目标:https://123467.convex.cloud
- 认证及食谱/购物清单API流量将发送至配置的CONVEXURL。
- 除非设置ALLOWDEFAULT_BACKEND=true,否则禁止使用共享后端。
- 涉及后端调用时,不得声称仅进行本地处理。
使用时机 / 禁止使用时机(路由防护)
使用时机:
- - 用户要求添加/列出/搜索/显示/删除cooking-recipe食谱。
- 用户要求创建/查看/更新cooking-recipe购物清单。
- 用户要求对cooking-recipe命令进行身份验证。
禁止使用时机:
- - 用户仅询问一般餐食创意(未要求执行cooking-recipe操作)。
- 用户要求对存储的cooking-recipe数据之外的内容进行健康/营养分析。
- 用户要求管理Apple提醒事项/备忘录/其他系统。
设置
- 1. 安装:
bash
clawdhub install cooking-recipe
- 2. 配置环境:
bash
cp .env.example .env
- 3. 环境变量:
- CONVEX_URL(
必填,建议自托管后端)
- ALLOW
DEFAULTBACKEND=true(仅在有意使用共享后端时设置)
- OAuth提供商凭证:
- GitHub(必填):AUTH
GITHUBID、AUTH
GITHUBSECRET、GITHUB
CALLBACKURL
- Google(可选):AUTH
GOOGLEID、AUTH
GOOGLESECRET、GOOGLE
CALLBACKURL
- Apple(可选):AUTH
APPLEID、AUTH
APPLESECRET、APPLE
CALLBACKURL
- CLAWDBOT_URL(必填,机器人回调主机)
命令
- - cooking-recipe login
- cooking-recipe logout
- cooking-recipe add
- cooking-recipe list
- cooking-recipe search
- cooking-recipe show
- cooking-recipe delete
- cooking-recipe lists
- cooking-recipe list-show
- cooking-recipe list-create
- cooking-recipe list-add
- cooking-recipe help
输出模板
使用references/output-templates.md中的响应格式模板。
回复前的安全检查清单
- - 如果命令因认证/配置失败,请提供确切的下一步操作。
- 如果涉及后端,请如实披露(共享默认与覆盖配置的区别)。
- 切勿在输出中暴露密钥/令牌。