hm-research

H&M research

Browse H&M's storefront categories, list or search products, and pull full product detail (colors, per-size stock, reviews) and nearby physical stores — all as normalized JSON from the Crawlora API, with no HTML scraping.

When to use this skill

• "What does X cost on H&M?" or "find X on H&M."
• "Browse H&M's {department} category" / "what's new in H&M's {department}?"
• "Search H&M for {keyword}."
• "Pull the colors/sizes/reviews for this H&M product."
• "Find the nearest H&M store to {zip/place}."

Setup (one-time)

• Get a free Crawlora API key (2,000 credits/mo, no card) at https://crawlora.net.
• export CRAWLORA_API_KEY=sk_your_key_here
• All requests: x-api-key: $CRAWLORA_API_KEY against https://api.crawlora.net/api/v1. Missing/invalid key → 401.

How it works

1. Browse categories — /hm/categories returns H&M's own storefront nav, department by department (optionally scoped with department). Entries include a verified category_id only where it's been confirmed against /hm/listing; not every category has one yet.
2. List or search — /hm/listing (category_id, an H&M category slug like ladies_newarrivals_all or men_newarrivals_all) or /hm/search (query, free-text) both return normalized products with pricing, images, colors, and per-size stock, paginated with page/page_size.
3. Detail — /hm/product/{product_id} (numeric id, taken from a listing/search result's id field or the productpage..html segment of its url) fetches every purchasable color with its own per-size price/availability, plus aggregate rating and real customer reviews with fit-feedback tags (e.g. "True to Size").
4. Stores — /hm/stores finds nearby physical H&M stores by search (zip/place, resolved to coordinates) or by lat/lng directly, within an optional radius_meters.

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

Calling the API

# Browse a category:
scripts/crawlora.sh /hm/listing category_id=ladies_newarrivals_all page=1 | jq '.'

# Free-text search:
scripts/crawlora.sh /hm/search query="linen shirt" | jq '.'

# Product detail (colors, sizes, reviews):
scripts/crawlora.sh /hm/product/1234567 | jq '{name,colors}'

# Nearby stores:
scripts/crawlora.sh /hm/stores search="98101" radius_meters=15000 | jq '.'

Raw curl fallback:

curl -fsS -H "x-api-key: $CRAWLORA_API_KEY" \
  "https://api.crawlora.net/api/v1/hm/search?query=linen+shirt" | jq '.'

Endpoint reference

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

Examples

• Category browse with detail: /hm/categories (optionally filtered by department) to find a category_id, then /hm/listing paginated to list products, then /hm/product/{product_id} for the picks worth a closer look.
• Keyword search with reviews: /hm/search for a query, then /hm/product/{product_id} on the top results to compare per-color pricing/stock and summarize customer reviews before recommending one.
• Local store check: /hm/stores with a search zip/place (or lat/lng) to list nearby stores before telling a shopper where to try something on.

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 H&M product/category/store pages; respect H&M's terms.
• Security: key lives in CRAWLORA_API_KEY only — never hardcode, query-param, or commit it.
• No category-tree lookup endpoint — /hm/listing's category_id values come from known H&M storefront slugs, not a discovery call; /hm/categories only exposes a verified category_id for the subset of nav entries this build has confirmed.
• Colors/sizes/reviews only come from product detail — /hm/listing and /hm/search carry one representative price and a per-color stock count each, not the full per-size breakdown or reviews.
• Pagination is real on /hm/listing and /hm/search — requesting a page past the last one returns an empty products array, not an error.