← 返回 Skills 市场
tobewin

China Recipe

作者 ToBeWin · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ✓ 安全检测通过
161
总下载
0
收藏
0
当前安装
3
版本数
在 OpenClaw 中安装
/install china-recipe
功能描述
中国菜谱大全。Use when user wants to find Chinese recipes, cooking instructions, ingredient lists, or meal planning. Supports regional cuisines, dietary preferences...
使用说明 (SKILL.md)

China Recipe

中国菜谱大全,支持各大菜系、食材查询、详细做法、营养信息。

Features

  • 🍜 菜系丰富: 川菜、粤菜、鲁菜、苏菜、浙菜、闽菜、湘菜、徽菜
  • 🥘 分类齐全: 荤菜、素菜、汤品、主食、甜点、凉菜
  • 📋 详细做法: 食材清单、步骤详解、烹饪技巧
  • 🔍 智能搜索: 按食材、菜名、口味搜索
  • 🥗 营养信息: 热量、蛋白质、脂肪、碳水
  • 🌍 多语言: 中英文支持

Trigger Conditions

  • "做什么菜" / "What to cook"
  • "给我一个菜谱" / "Give me a recipe"
  • "红烧肉怎么做" / "How to make braised pork"
  • "今天吃什么" / "What to eat today"
  • "有什么简单的菜" / "Easy recipes"
  • "china-recipe"

How It Works (工作原理)

数据来源

本skill不维护菜谱数据库,而是利用LLM的烹饪知识动态生成菜谱:

  1. 用户请求菜谱(如"红烧肉怎么做")
  2. Agent根据LLM知识生成详细菜谱
  3. 输出标准化格式的菜谱报告

优势

  • 无限扩展 - LLM知道几乎所有菜谱
  • 实时生成 - 根据需求动态生成
  • 无需维护 - 无需维护数据库
  • 个性化 - 可根据口味、人数调整

Step 1: 识别用户需求

用户输入 → 分析需求:

"红烧肉怎么做"
  → 检测:查询具体菜谱
  → 生成:红烧肉详细做法

"有什么简单的菜"
  → 检测:推荐简单菜
  → 生成:3-5道简单菜谱

"川菜推荐"
  → 检测:推荐菜系
  → 生成:川菜代表菜

"今天吃什么"
  → 检测:随机推荐
  → 生成:2-3道随机菜谱

"用鸡胸肉做什么菜"
  → 检测:按食材推荐
  → 生成:鸡胸肉菜谱

Step 2: 返回菜谱详情

菜谱详情格式

┌──────────────────────────────────────────────┐
│  🍳 红烧肉                                     │
│  川菜 | 难度:⭐⭐⭐ | 时间:90分钟 | 4人份  │
└──────────────────────────────────────────────┘

📝 食材清单
├─ 五花肉 500g
├─ 生姜 3片
├─ 大葱 2段
├─ 八角 2个
├─ 桂皮 1小块
├─ 冰糖 30g
├─ 生抽 2勺
├─ 老抽 1勺
└─ 料酒 2勺

👨‍🍳 做法步骤
1. 五花肉切块,冷水下锅焯水
2. 锅中放油,加冰糖炒糖色
3. 放入肉块翻炒上色
4. 加入调料和热水,大火烧开
5. 转小火炖60分钟
6. 大火收汁即可

💡 烹饪技巧
├─ 选肉:肥瘦相间的五花肉最佳
├─ 炒糖色:小火慢炒,防止焦苦
└─ 炖煮:小火慢炖,肉质软烂

🔥 营养信息
├─ 热量:450 kcal/100g
├─ 蛋白质:15g
├─ 脂肪:40g
└─ 碳水:5g

Search Logic (搜索逻辑)

RECIPES = {
    "红烧肉": {
        "cuisine": "川菜",
        "difficulty": 3,
        "time": 90,
        "servings": 4,
        "taste": "咸甜",
        "ingredients": [
            "五花肉 500g",
            "生姜 3片",
            "大葱 2段",
            "八角 2个",
            "桂皮 1小块",
            "冰糖 30g",
            "生抽 2勺",
            "老抽 1勺",
            "料酒 2勺"
        ],
        "steps": [
            "五花肉切块,冷水下锅焯水",
            "锅中放油,加冰糖炒糖色",
            "放入肉块翻炒上色",
            "加入调料和热水,大火烧开",
            "转小火炖60分钟",
            "大火收汁即可"
        ],
        "tips": [
            "选肉:肥瘦相间的五花肉最佳",
            "炒糖色:小火慢炒,防止焦苦",
            "炖煮:小火慢炖,肉质软烂"
        ],
        "nutrition": {
            "calories": 450,
            "protein": 15,
            "fat": 40,
            "carbs": 5
        }
    },
    "番茄炒蛋": {
        "cuisine": "家常菜",
        "difficulty": 1,
        "time": 10,
        "servings": 2,
        "taste": "酸甜",
        "ingredients": [
            "番茄 2个",
            "鸡蛋 3个",
            "葱花 适量",
            "盐 适量",
            "糖 1勺",
            "油 适量"
        ],
        "steps": [
            "番茄切块,鸡蛋打散",
            "锅中倒油,炒散鸡蛋盛出",
            "锅中留油,炒番茄出汁",
            "加入盐和糖调味",
            "放入鸡蛋翻炒均匀",
            "撒葱花出锅"
        ],
        "tips": [
            "番茄要选熟透的,出汁多",
            "鸡蛋要炒嫩一点",
            "加糖可以提鲜"
        ],
        "nutrition": {
            "calories": 150,
            "protein": 8,
            "fat": 10,
            "carbs": 8
        }
    }
}

def search_recipe(query):
    """智能搜索菜谱"""
    
    # 按菜名搜索
    for name, recipe in RECIPES.items():
        if query in name or name in query:
            return {name: recipe}
    
    # 按食材搜索
    results = {}
    for name, recipe in RECIPES.items():
        for ingredient in recipe['ingredients']:
            if query in ingredient:
                results[name] = recipe
                break
    
    # 按难度搜索
    if '简单' in query:
        for name, recipe in RECIPES.items():
            if recipe['difficulty'] \x3C= 2:
                results[name] = recipe
    
    # 按菜系搜索
    for name, recipe in RECIPES.items():
        if query in recipe.get('cuisine', ''):
            results[name] = recipe
    
    return results or RECIPES

Notes

  • 菜谱数据内置,无需外部API
  • 支持多种搜索方式
  • 支持中英文输出
  • 完全本地运行
安全使用建议
This skill appears coherent and limited in scope: it needs only python3, contains local recipe data/search logic, and asks for no credentials or installs. Before installing, you may (1) inspect the raw SKILL.md for any hidden/invisible characters (the scanner flagged unicode-control-chars), (2) confirm whether you prefer purely LLM-generated recipes or the embedded local recipes (the doc mentions both), and (3) run it in a least-privileged environment or sandbox if you want to observe behavior first. Because the skill can be invoked autonomously by the agent (normal default), review agent autonomy settings if you prefer to control when skills run.
功能分析
Type: OpenClaw Skill Name: china-recipe Version: 1.0.2 The 'china-recipe' skill is a straightforward tool for providing Chinese cooking instructions and recipes. The content in SKILL.md consists of standard operational instructions for the AI agent and a simple Python logic snippet for searching a hardcoded recipe dictionary; it contains no network requests, file system access, or suspicious execution patterns.
能力评估
Purpose & Capability
Name/description (China Recipe) match the skill contents. Requesting python3 is proportional because SKILL.md includes a small Python search snippet. Minor inconsistency: the doc says the skill uses an LLM to dynamically generate recipes but also embeds a local RECIPES dictionary and states "菜谱数据内置" — this is plausible (hybrid behavior) but worth noting.
Instruction Scope
SKILL.md instructs the agent to generate or return recipes and includes a local search function. It does not ask the agent to read arbitrary files, access environment variables, call external endpoints, or exfiltrate data. All runtime behavior described is limited to recipe generation/search and formatting.
Install Mechanism
There is no install spec (instruction-only). That is low risk. Requiring python3 is reasonable given the included Python snippet; nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The absence of secrets or unrelated service tokens is appropriate for a local recipe helper.
Persistence & Privilege
always is false and the skill does not request persistent/system-wide configuration or elevated privileges. It does not modify other skills or agent-wide settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install china-recipe
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /china-recipe 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.2
修复依赖声明:移除未使用的curl
v1.0.1
改为LLM驱动:利用AI知识动态生成菜谱,无限扩展,无需维护数据库
v1.0.0
中国菜谱大全:川粤鲁苏浙闽湘徽八大菜系,食材/菜名/口味搜索,详细做法+营养信息
元数据
Slug china-recipe
版本 1.0.2
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 3
常见问题

China Recipe 是什么?

中国菜谱大全。Use when user wants to find Chinese recipes, cooking instructions, ingredient lists, or meal planning. Supports regional cuisines, dietary preferences... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 161 次。

如何安装 China Recipe?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install china-recipe」即可一键安装,无需额外配置。

China Recipe 是免费的吗?

是的,China Recipe 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

China Recipe 支持哪些平台?

China Recipe 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 China Recipe?

由 ToBeWin(@tobewin)开发并维护,当前版本 v1.0.2。

💬 留言讨论