From find-apartments
Use when scraping WinWin rental listings for a specific city and room count — uses web_search to find listings then web_fetch on each individual listing page to extract full details
How this skill is triggered — by the user, by Claude, or both
Slash command
/find-apartments:search-winwinThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Scrapes rental apartment listings from winwin.co.il for a given city and minimum room count.
Scrapes rental apartment listings from winwin.co.il for a given city and minimum room count.
You will receive search parameters:
city — city name in Hebrew (e.g., "הוד השרון")min_rooms — minimum number of rooms (e.g., 5)max_price — optional maximum monthly rent in NISUse WebSearch with domain filter:
WebSearch: "דירה להשכרה {city} {min_rooms} חדרים site:winwin.co.il"
Use WebFetch on the WinWin search results page to extract individual listing URLs. Look for URLs matching patterns like:
https://www.winwin.co.il/RealEstate/ForRent/Ads/RealEstateAds,XXXXXXX.aspx
For each listing URL found, fetch full details:
WebFetch URL: <listing URL>
Prompt: "Extract ALL details about this rental listing: full address, price, rooms, floor, size in sqm, amenities (parking, elevator, mamad, balcony, AC, storage, pets), available date, contact info, and image URLs."
Return each listing in the normalized format:
{
"url": "https://www.winwin.co.il/RealEstate/ForRent/Ads/RealEstateAds,XXXXXXX.aspx",
"source": "WinWin",
"city": "{city}",
"address": "extracted address",
"neighborhood": "extracted neighborhood or empty string",
"rooms": 5,
"floor": "3/8",
"size_sqm": 150,
"price": 8000,
"mamad": true,
"parking": "1",
"elevator": true,
"ac": true,
"balcony": true,
"pets": null,
"available_from": "",
"contact": null,
"images": []
}
Return the complete array of normalized listing objects. Report count found and count successfully fetched.
npx claudepluginhub cxt9/find-apartments --plugin find-apartmentsCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.