BrowserOS compare-prices

Search for a product across multiple retailers in parallel, save pricing data to disk, and produce an HTML report with the best deals and direct product links. Use when the user asks to compare prices, find the best deal, or check prices across stores.

install

source · Clone the upstream repo

git clone https://github.com/browseros-ai/BrowserOS

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/browseros-ai/BrowserOS "$T" && mkdir -p ~/.claude/skills && cp -r "$T/packages/browseros-agent/apps/server/src/skills/defaults/compare-prices" ~/.claude/skills/browseros-ai-browseros-compare-prices && rm -rf "$T"

manifest: packages/browseros-agent/apps/server/src/skills/defaults/compare-prices/SKILL.md

source content

Compare Prices

Search for a product across retailers in parallel using a hidden window, save pricing data incrementally to disk, and deliver a clean HTML comparison report with direct links to every product page.

When to Apply

Activate when the user asks to compare prices for a product, find the cheapest option, check if a price is good, or shop across multiple stores.

Workflow

Phase 1 — Clarify

Confirm with the user before searching:

Product name — exact model, variant, size, or color if applicable
Retailer preferences — any stores to include or exclude
Region / currency — defaults to user's locale

Phase 2 — Set Up & Search

Step Tool Detail

Create output directory

evaluate_script

Create

compare-<product-slug>/

in your working directory with a

raw/

subfolder

Open hidden window

create_hidden_window

Dedicated workspace — keeps the user's browsing undisturbed

Open parallel tabs

new_hidden_page

Open up to 10 tabs concurrently, one per retailer/search

Default search targets (adjust based on product type and user's region):

Tab	Target
1	Google Shopping — `https://www.google.com/search?tbm=shop&q=<product>`
2	Amazon — `https://www.amazon.com/s?k=<product>`
3	Walmart — `https://www.walmart.com/search?q=<product>`
4	Best Buy — `https://www.bestbuy.com/site/searchpage.jsp?st=<product>`
5	Target — `https://www.target.com/s?searchTerm=<product>`
6	eBay — `https://www.ebay.com/sch/i.html?_nkw=<product>`
7–10	Additional retailers relevant to the product category (Newegg for tech, Home Depot for tools, etc.)

Phase 3 — Extract & Save

For each tab, extract pricing data and save immediately:

Step	Tool	Detail
Navigate	`navigate_page`	Go to the search URL
Read results	`get_page_content`	Extract the search results page as markdown
Find best match	`navigate_page`	Click through to the most relevant product listing
Extract pricing	`get_page_content`	Pull the product page content — price, availability, shipping, seller
Save raw data	`evaluate_script`	Write to `raw/<retailer>.json` with all extracted fields (see format below)
Close tab	`close_page`	Free the tab after saving

Never hold all retailer data in memory. Save each retailer's data to its own file immediately after extraction.

Raw Data Format (

raw/<retailer>.json

)

{
  "retailer": "Amazon",
  "product_name": "Product Title as Listed",
  "product_url": "https://www.amazon.com/dp/...",
  "price": 299.99,
  "original_price": 349.99,
  "currency": "USD",
  "shipping": "Free",
  "availability": "In Stock",
  "seller": "Amazon.com",
  "condition": "New",
  "rating": "4.5/5",
  "notes": "Prime eligible",
  "extracted_at": "2025-03-11T10:30:00Z"
}

Phase 4 — HTML Report

After all retailers are processed, read the saved

raw/*.json

files and generate a self-contained

report.html

Requirement	Detail
Theme	Light background ( `#ffffff` ), clean sans-serif typography, generous whitespace
Header	Product name, search date, number of retailers checked
Best Deal banner	Highlighted card at the top showing the lowest total price with a direct link to the product page
Comparison table	All retailers sorted by total price (lowest first) with columns: Retailer, Price, Shipping, Total, Stock, Seller, Rating, Link
Product links	Every retailer name and a "View Deal" button must be a clickable `<a href>` linking to the actual product page URL
Price highlights	Lowest price in green, highest in muted gray. Show discount percentage if original price differs.
Self-contained	All styles in a `<style>` block — no external CSS or JS
Responsive	Readable on desktop and mobile
Footer	"Generated by BrowserOS Compare Prices" with date

Use

evaluate_script

to write

report.html

to the output directory.

Phase 5 — Open & Notify

Step	Tool	Detail
Close hidden window	`close_window`	Clean up the research workspace
Open report	`new_page`	Open `file://<path>/report.html` in the user's active window
Notify user	—	Tell the user the comparison is complete, highlight the best deal, and provide the report path

Tool Reference

Category	Tools Used
Window management	`create_hidden_window` , `close_window`
Tab management	`new_hidden_page` , `close_page` , `new_page`
Navigation	`navigate_page`
Content extraction	`get_page_content`
Data & file I/O	`evaluate_script`

Tips

Always compare total price (product + shipping), not just the listed price.
Note the seller — marketplace third-party sellers may have different return policies than the retailer itself.
Mention membership discounts (Prime, Walmart+) as a note, not as the default price.
If the product has variants (sizes, colors), ensure every retailer is quoting the same variant.
If a retailer blocks scraping or returns no results, skip it and note the gap in the report.
For used/refurbished listings, separate them from new-condition results.