milwaukeetool.com.au — AI Site Grade

Analysis

The llms.txt endpoint returns a 410 Gone — a full HTML page with a Salesforce Commerce Cloud JS shell, not a simple 404 — meaning AI crawlers hitting it get a 381KB blob of JavaScript rather than a clean content map.

Crawler Access

All major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended) receive 200 OK with full HTML content from the homepage. No UA-based blocking exists. The site runs on Salesforce Commerce Cloud (Demandware) behind Cloudflare with no security headers (no HSTS, no CSP, no X-Frame-Options). The robots.txt uses a single User-agent: * rule with no AI-bot-specific directives — GPTBot, ClaudeBot, and others are governed by the same catch-all that disallows search result pages, filter parameters, and pagination URLs. The sitemap index at /sitemap_index.xml lists 5,070+ product URLs across two sub-sitemaps, but list_known_urls only discovered 25 URLs from the homepage nav — the sitemap is the only complete inventory.

Cold-Knowledge Gap

The LLM prior knows Milwaukee Tool Australia as a professional-grade power tool brand with M18/M12/MX FUEL platforms, PACKOUT storage, and a Trade Loyalty program. The site itself confirms these but never mentions PACKOUT by name on any fetched page — the word does not appear in homepage text, product pages, or the ONE-KEY page. The cold model also references a "Trade Loyalty" program, yet the site's robots.txt blocks */Product-Show (the product detail endpoint pattern), and the redemption page (/redemptions/) describes promotional offers but does not use the phrase "Trade Loyalty." The model knows about the 1924 Hole Shooter invention — a historical fact absent from the site's homepage and product pages.

Schema Posture

The homepage includes a rich Organization + WebSite + WebPage JSON-LD graph with social profiles, logo, and search action. However, product pages lack any Product schema — the M18FCSDC1650 circular saw page returns zero JSON-LD types. The recall page and the ONE-KEY page also have no structured data. The recall page does include a well-structured FAQ section (collapsible Q&A) but no FAQPage schema to mark it up for search engines or AI answer extraction.

External Signals

The site links to a Product Safety Australia recall PDF for the chainsaw recall, but web searches return zero indexed news articles, Reddit threads, or press mentions about Milwaukee Tool Australia specifically. The learn.milwaukeetool.com.au subdomain redirects to a login wall with zero visible text — a JS-rendered gated portal invisible to crawlers. The catalogue.milwaukeetool.com.au subdomain hosts a flipbook catalogue. The brand has active social channels (Facebook, Instagram, TikTok, YouTube, LinkedIn) but no blog, no articles, and no editorial content on the main domain — the site is a pure product catalog with no informational content for AI engines to cite.

Findings

1. llms.txt returns 410 Gone — The endpoint exists but serves a full HTML page (381KB) with a JS shell, not a plain text content map. This is worse than a 404 because crawlers waste bandwidth parsing a rendered page that yields nothing.

2. No AI-bot directives in robots.txt — GPTBot, ClaudeBot, PerplexityBot, and Google-Extended all follow the same * rules. No explicit allowance or disallowance exists for any AI crawler.

3. Product pages lack Product schema — The M18FCSDC1650 circular saw page has zero JSON-LD. No Product, Offer, AggregateRating, or BreadcrumbList schema on any product page sampled.

4. FAQ page lacks FAQPage schema — The chainsaw recall page has 11 collapsible Q&A items but no FAQPage markup, making them invisible to AI answer extraction.

5. PACKOUT never mentioned on-site — Despite being a flagship product system globally, the word "PACKOUT" does not appear in any fetched page text or headings.

6. No blog or informational content — The domain has zero articles, guides, or trade-specific content. The learn. subdomain is a login-gated portal with no crawlable content.

7. Sitemap contains 5,070+ URLs but homepage nav only exposes ~25 — The sitemap is the only complete URL inventory. Category pages and discontinued product pages are in the sitemap but not linked from the main navigation.

8. No security headers — Missing HSTS, CSP, X-Frame-Options, X-Content-Type-Options, and Referrer-Policy headers. The site sets cache-control: no-cache, no-store, must-revalidate on the homepage.

Findings

1. llms.txt endpoint returns 410 Gone with JS shell High

   The llms.txt endpoint returns a 410 Gone status with a full HTML page containing a JavaScript shell, wasting crawler bandwidth and providing no content map.

   What to change: Replace the llms.txt endpoint with a plain text file listing key URLs and content summaries, or remove the endpoint entirely to avoid wasted crawler resources.

2. No AI-bot-specific directives in robots.txt Medium

   The robots.txt uses a single User-agent: * rule with no explicit allowances or disallowances for AI crawlers like GPTBot, ClaudeBot, or Google-Extended.

   What to change: Add explicit directives for AI crawlers, such as allowing GPTBot and ClaudeBot to access product pages while disallowing search result parameters.

3. Product pages lack Product schema markup High

   The M18FCSDC1650 circular saw page and other product pages contain zero JSON-LD structured data, missing Product, Offer, and BreadcrumbList schemas.

   What to change: Add Product schema with name, description, SKU, price, and availability to all product pages.

4. FAQ section on recall page lacks FAQPage schema Medium

   The chainsaw recall page has 11 collapsible Q&A items but no FAQPage structured data, preventing AI answer extraction.

   What to change: Add FAQPage schema markup to the Q&A section on the recall page.

5. PACKOUT product system not mentioned on site Medium

   Despite being a flagship product system globally, the word 'PACKOUT' does not appear in any fetched page text or headings.

   What to change: Add PACKOUT content to relevant product pages and category pages to align with brand recognition.

6. No blog or informational content on domain Medium

   The domain has zero articles, guides, or trade-specific content; the learn subdomain is a login-gated portal with no crawlable content.

   What to change: Create informational content such as buying guides, how-to articles, or trade tips to provide AI-citable material.

7. Sitemap contains 5,070+ URLs but homepage nav exposes only ~25 Low

   The sitemap lists over 5,000 product URLs, but the homepage navigation only links to about 25 pages, leaving many products undiscoverable via navigation.

   What to change: Improve internal linking to surface more product categories and pages from the main navigation.

8. Missing security headers on homepage Low

   The homepage lacks HSTS, CSP, X-Frame-Options, X-Content-Type-Options, and Referrer-Policy headers, though this does not directly impact AI visibility.

   What to change: Add security headers such as HSTS, CSP, and X-Frame-Options to improve overall site security posture.

What's working

• All major AI crawlers receive 200 OK with full HTML — GPTBot, ClaudeBot, PerplexityBot, and others are not blocked and receive full HTML content from the homepage.
• Homepage includes rich Organization and WebSite JSON-LD — The homepage has a comprehensive JSON-LD graph with Organization, WebSite, and WebPage schemas, including social profiles and search action.
• Sitemap index with 5,070+ product URLs is accessible — The sitemap index at /sitemap_index.xml lists over 5,000 product URLs, providing a complete inventory for crawlers.
• Recall page has well-structured FAQ content — The chainsaw recall page includes 11 collapsible Q&A items that are human-readable and well-organized.