Wish Research

Wish research

Browse Wish's category tree, search its product catalog, and pull product detail, related items, and customer reviews — all as normalized JSON from the Crawlora API, with no HTML scraping.

When to use this skill

• "What does X cost on Wish?" or "find X on Wish."
• "Browse Wish's categories" / "what's under [department] on Wish?"
• "Pull the detail/price/rating for this Wish product."
• "What are the reviews like for this Wish product?"
• "Find similar or related items to this Wish product."

Setup (one-time)

• Get a free Crawlora API key (2,000 credits/mo, no card) at https://crawlora.net.
• Set CRAWLORA_API_KEY in the environment before running the helper.
• The helper reads CRAWLORA_API_KEY from the environment and sends requests to https://api.crawlora.net/api/v1. Missing/invalid key → 401.

How it works

1. Browse — /wish/categories returns Wish's own top navigation/category tree plus each category's nested filter groups where present. It's a static, site-wide taxonomy — no input, doesn't vary by search term.
2. Search — /wish/search (query) searches the product catalog by keyword, offset-paginated, returning price, currency, rating, review count, and merchant id per product. Use /wish/suggest (query) first for typeahead search-term ideas if the user's query is vague.
3. Detail — /wish/product/{id} (id is the 24-character hex product_id from a search result or product URL) fetches full product detail: name, description, sold-out state, aggregate rating, images, and every purchasable variation with its own price, currency, inventory, and merchant.
4. Related items — /wish/product/{id}/related returns similar-item rails for that product (a general-similarity rail and a faster-shipping subset where eligible).
5. Reviews — /wish/product/{id}/reviews returns the product's normalized customer reviews.
6. Compare the JSON fields (price, rating, review count, merchant) across products or reviews and answer.

Full endpoint list, methods, and params: reference/endpoints.md.

Calling the API

# Search Wish (GET, key=value params):
scripts/crawlora.sh /wish/search query="wireless earbuds" count=30 | jq '.'

# Search-term suggestions:
scripts/crawlora.sh /wish/suggest query="wireless ear" | jq '.'

# Product detail:
scripts/crawlora.sh /wish/product/5f8a1c2e9b3d4a001f7e6c21 | jq '.data'

# Reviews:
scripts/crawlora.sh /wish/product/5f8a1c2e9b3d4a001f7e6c21/reviews count=50 | jq '.'

Use scripts/crawlora.sh for all requests; it keeps the API key out of command-line arguments.

Endpoint reference

See reference/endpoints.md for every Wish endpoint this skill uses (method, path, params, description).

Examples

• Search and compare: /wish/search for a keyword, compare price, rating, and review count across results to shortlist candidates.
• Product due diligence: /wish/product/{id} for price/rating/variations, then /wish/product/{id}/reviews to summarize what buyers say before recommending it.
• Find similar items: /wish/product/{id}/related to pull a product's similar-item rails when the exact item is sold out or the user wants alternatives.

Notes & limits

• Credits / pay-on-success: billed only on 2xx; free tier 2,000 credits/mo. Key at https://crawlora.net.
• Public data only — public Wish product/category pages; respect Wish's terms.
• Security: key lives in CRAWLORA_API_KEY only — never hardcode, query-param, or commit it.
• Product identity is id (a 24-character hex product_id), not a SKU or ASIN — take it from a search result's product_id field or a product page's URL; /wish/product, /wish/product/related, and /wish/product/reviews all key off it. An unrecognized id returns 404 on /wish/product but a normal empty result on related/reviews.
• /wish/search pagination is offset-based, not page-based: offset must be an exact multiple of count, up to 3 * count (so at most 4 pages reachable per query) — pass a larger count (max 70) rather than paging deep.
• /wish/product/{id}/reviews has no offset/cursor — the upstream source doesn't support one, so re-request with a larger count (max 200) to get more reviews, not a page/offset param.
• A query with no matches, or a product with zero reviews, returns a normal empty result rather than an error.