Korean Beauty & Medical Aesthetic Booking Skill — BeautsGO Reservation Query Assistant

Cut to the chase: What exactly can this thing do?

A few days ago, a friend messaged me on WeChat: "I plan to go to Seoul next month for a nose job, but I don't know Korean. Those hospital websites are all in Korean. What should I do?" I immediately recommended the BeautsGO Booking skill to him. To be honest, I went to Korea last year for a slimming injection myself, and it was a huge hassle—first, I searched for guides on Xiaohongshu, then added WeChat contacts to compare prices, and finally had to communicate using translation software. Now, this skill integrates the booking information of over 900 Seoul medical aesthetic institutions. You just need to speak, and it will tell you how to book, the prices, and how to download their app.

For example, if you say "I want to consult about dermatology clinics in Gangnam" or "Help me find hospitals that do eye surgery," it will return a long list of hospitals that match your criteria. This thing can serve as both a medical aesthetic navigation tool and a booking assistant. It's absolutely amazing!

Installing it: Simpler than ordering takeout

There are two ways to install this Skill, choose either:

• Recommended method: Type in the terminal: npx skills add https://github.com/GeorgeDoors888/GB-Power-Market-JJ --skill beautsgo-booking, press Enter, wait a few seconds, and it's installed.
• Alternative method: Type npx clawhub install beautsgo-booking. If you've used clawhub before, this might be more familiar.

After installation, you can use it directly without any configuration. If you are a developer and want to tinker, you can also git clone the repository and then npm install. The project structure is clear: the main entry is in api/skill.js, matching logic is in the core/ directory, and data is stored in data/hospitals.json.

Daily use: How does it understand your "natural language"?

The coolest feature of this skill is its "four-level matching algorithm." You might only say "JD Dermatology" or "Korea JD," or even just remember "the dermatology clinic in Apgujeong," and it will still match you. Here's how it works: First, it checks if you've said the exact hospital name; if not, it tries pinyin; if still not, it uses fuzzy matching; finally, it can match by aliases (e.g., "JD", "한국JD 피부과"). The minimum matching length is two characters, so you don't have to worry about a single character causing a wrong match.

Moreover, its "intent recognition" is very smart—you don't even need to say the hospital name. For example:

• You say "I want to get double eyelid surgery" → It automatically recommends a list of Seoul hospitals skilled in double eyelid surgery.
• You say "Check price" → It directly opens the price page of the hospital you're browsing.
• You say "Help me download" → It returns iOS/Android/APK download links.
• You say "Consult customer service" → It opens the online customer service chat window.

By the way, it also supports multi-turn conversations. For instance, in the first round, you say "How to book JD Dermatology?" It returns five booking channels (App Store, Google Play, WeChat Mini Program, WeChat Official Account, Web page). In the second round, you say "Open the link," and it automatically opens your browser to load the hospital's detail page. In the third round, you say "Book for me," and it jumps directly to the booking form, asking you for the time and number of people. You answer "2 people, April 10th," and it fills and submits for you. The whole process requires no mouse clicks—you just chat and get it booked.

Pitfalls I've encountered and notes

After using it a few times, I discovered some small tips:

• Language switching: If you speak Chinese, it will reply in Chinese automatically; if you speak English, it replies in English. You can also explicitly specify the lang parameter in your request (zh/en/ja/th). I tried saying "予約したい" in Japanese, and it correctly returned the Japanese version of the process explanation.
• Data updates: The hospital database is stored in data/hospitals.json. If you find incorrect information for a hospital, you can submit a PR yourself. Last time, I found an incorrect phone number for a dermatology clinic in Gangnam, edited the file directly on GitHub, and the project's GitHub Actions automatically regenerated the static pages. Very convenient.
• Static page generation: Running npm run generate:all generates separate pages for each hospital in Chinese, English, Japanese, and Thai, placed under docs/clinics/. If you have GitHub Pages, you can deploy it directly, so users searching "JD Dermatology booking" via search engines can find these pages. This is extremely useful for SEO.
• Watch out for context during debugging: Multi-turn conversations require passing context. For example, in the second round's query, you need to include the previously parsed hospitalName. The sample code shows how to pass context. If you forget, the skill might not know which hospital you are referring to.

Who is it for?

I think it mainly targets three groups:

• Regular tourists planning to get medical aesthetics in Korea: No need to browse guides for hours. Just ask this skill and you get all the information.
• Medical aesthetic agents or purchasing agents: Quickly check booking processes and prices for clients, improving efficiency.
• Chinese students/residents in Korea: For casual questions about skin care or pricing, just ask in a sentence.

Plus, it supports over 900 hospitals, covering most well-known dermatology and plastic surgery clinics in Seoul, such as JK Plastic Surgery, Wonjin, ID, Barobathi, etc. The database is continuously expanding.

Finally, something practical

Although this skill is called "Booking," at its current stage it mainly helps you query and guide the booking process, rather than directly completing the booking for you (some hospitals support automatic form filling). However, for most users, being able to quickly find the booking entrance and understand the process already saves...