skill / Product & Materials Research

Product Spec Bulk Fetch

A Claude Code skill that takes a list of product page URLs and extracts structured furniture specification data — names, dimensions, materials, pricing, and images — into a standardized FF&E schedule.

procurement data-lookup
Install ↓ View on GitHub →
New to skills? Read the cheat sheet for architects and designers →

What it does

Type /product-spec-bulk-fetch in Claude Code and paste a list of product URLs. The skill fetches each page, extracts structured FF&E data using AI, and compiles a clean schedule ready for procurement, spec books, or import into Canoa.

Works with any furniture vendor site — Herman Miller, Steelcase, IKEA, RH, DWR, HAY, 1stDibs, local vendors in any language.

Install

Claude Desktop:

  1. Open Customize → Browse plugins
  2. Click + → Add marketplace from GitHub
  3. Enter AlpacaLabsLLC/skills-for-architects
  4. Install the Product & Materials Research plugin

Claude Code (terminal):

claude install github:AlpacaLabsLLC/skills-for-architects/05-materials-research

Usage

/product-spec-bulk-fetch

Then provide URLs — paste them inline, point to a .txt file, or give a Google Sheet ID with a column of URLs.

/product-spec-bulk-fetch

https://www.hermanmiller.com/products/seating/lounge-chairs/eames-lounge-chair/
https://www.steelcase.com/products/collaborative-chairs/gesture/
https://www.ikea.pr/puertorico/es/pd/vardagen-vaso-art-70313106

The skill fetches pages in parallel (up to 5 at a time) and extracts 15 fields per product.

Output schema

Every product is extracted into the Canoa Clipper FF&E format:

FieldExample
Product NameEames Lounge Chair
BrandHerman Miller
CollectionEames
CategorySeating
W / D / H32.75 / 32.5 / 33.5
Unitin
MaterialsMolded plywood, leather
Colors/FinishesWalnut / Black Leather
Price5695.00
CurrencyUSD
Lead Time8-10 weeks
URL(source page)
Image URL(hero image)

Output formats

After extraction, choose where to save:

  • Local CSV — ~/Documents/Work-Docs/ffe-fetch-YYYY-MM-DD.csv
  • Google Sheet — writes rows directly to a specified spreadsheet
  • Markdown table — stays in the conversation for quick review

How it handles failures

Not every page cooperates. The skill groups results into three buckets:

  • Successful — all key fields extracted
  • Partial — some fields missing (price on trade sites, dimensions on lifestyle pages). Row still included with blanks.
  • Failed — non-product page, login wall, or fetch error. Logged and skipped.

After the batch: Fetched: 12/15 successful, 2 partial, 1 failed

Pairs with

Use /product-spec-bulk-cleanup after fetching to normalize casing, translate Spanish fields, standardize dimensions, and deduplicate. The two skills chain: fetch → cleanup → spec-ready schedule.