food-travel — Eat-First Travel Planner

One-liner: Input a dish, a craving, or a city — get a complete travel plan built around eating.

This skill solves the full "eat → where → go → stay → route" chain for food lovers.

Scenario Detection

Identify which scenario the user wants, then follow the corresponding workflow:

If unclear, ask the user to clarify.

Scenario A: Pick a Destination for Food

Input: a dish, cuisine, or flavor preference
Output: best city recommendation + food list + travel logistics

Steps

1. 1. Web search: "{dish/cuisine} 最正宗 去哪个城市吃" to identify the top 2-3 cities.
2. For each city, web search: "{city} 必吃 {dish} 餐厅推荐" to get restaurant data.
3. Search flights (if user provides origin):

   flyai search-flight --origin "{origin}" --destination "{city}" --dep-date {date}
   

1. 4. Search hotels:

Output format

CODEBLOCK2

Scenario B: City Food Map

Input: a city name
Output: meal-by-meal restaurant map organized by time of day

Steps

1. 1. Web search: "{city} 必吃餐厅推荐" + "{city} 特色小吃 推荐" + "{city} 夜宵 推荐".
2. Organize results into 4 time slots: 早餐, 午餐, 晚餐, 夜宵/下午茶.
3. keyword-search supplement:

   flyai keyword-search --query "{city} 美食券 餐厅"
   

Output format

CODEBLOCK4

Scenario C: Full Food-Driven Itinerary

Input: city + duration (e.g. "3天吃遍西安")
Output: day-by-day schedule with every meal planned + attractions between meals + transport + hotel

Steps

1. 1. Web search: "{city} {N}天美食攻略" + "{city} 必吃餐厅推荐".
2. Search hotels:

   flyai search-hotel --dest-name "{city}" --check-in-date {date} --check-out-date {date}
   

1. 3. Search flights (if origin provided):

   flyai search-flight --origin "{origin}" --destination "{city}" --dep-date {date}
   

1. 4. Search attractions to fill between-meal time:

   flyai search-poi --city-name "{city}"
   

1. 5. Organize into a day-by-day plan where every meal is the anchor.

Output format

CODEBLOCK8

General Rules

• - Food comes first — every itinerary section starts with a meal, attractions fill the gaps.
• Web search for restaurants — flyai has no restaurant database; always use web search for dining data.
• flyai for logistics — use search-flight, search-hotel, search-poi, keyword-search for transport, accommodation, attractions, and bookable dining products.
• Always include booking links — for every flight, hotel, and POI result, show [Click to book]({url}).
• Always include images — show ![]({picUrl}) or ![]({mainPic}) when available.
• Practical details — include price, address, opening hours when available.
• Source attribution — "餐厅数据来自网络搜索，机票酒店来自 fly.ai 实时结果。"