# How to Pass the New 'Agentic Browsing' Audit in Shopify PageSpeed > Learn how to pass Shopify PageSpeed Insights' new Agentic Browsing audit. Optimize your theme for AI agents using /agents.md and clean HTML structure. - Source: https://seo-accelerator.nicheflash.com/blogs/shopify-agentic-browsing-page-speed-insights-audit - Publisher: Shopify SEO Accelerator - Published: 2026-08-08 - Updated: 2026-08-08 ### Key takeaways - The Agentic Browsing audit is an experimental category in PageSpeed Insights that evaluates a site's readiness for autonomous AI shopping assistants rather than solely human users. - The audit provides a pass ratio (e.g., 3/3) based on three criteria: clean accessibility trees, layout stability, and the existence of a canonical `/agents.md` file. - Shopify merchants must implement a custom `templates/agents.md.liquid` file to satisfy the AI discovery requirement and structure data for automated agents. - While not a direct ranking factor, optimizing for Agentic Browsing improves technical SEO health and prepares stores for the Universal Commerce Protocol. ## What is the Agentic Browsing audit in PageSpeed Insights? The Agentic Browsing audit is an experimental category added to Lighthouse 13.3, which was released on May 21, 2026, and it measures whether a website is structured effectively for autonomous machine interaction rather than just human browsing. This score appears alongside Core Web Vitals in Google PageSpeed Insights and assesses your store's ability to support AI shopping assistants that operate like bots, navigating product pages and adding items to carts without human intervention. The Chrome for Developers documentation states that "The Agentic Browsing category evaluates how well your site is constructed for machine interaction through a set of deterministic audits." As noted in recent updates, these tests are designed to capture structural signals that traditional performance metrics often overlook. ## Which technical criteria does the audit verify? The audit verifies three primary criteria—clean accessibility trees, layout stability, and the existence of a canonical AI discovery file—to determine if a store passes the machine-interaction test. Unlike traditional performance metrics that generate a weighted score from 0-100, the Agentic Browsing audit provides a pass ratio, such as 3/3 audits passed. This binary outcome emphasizes strict adherence to specific structural standards required by automated agents. The criteria checked as of August 2026 include: - **Clean Accessibility Tree:** A clean accessibility tree ensures all page elements are programmatically exposed via ARIA labels and semantic HTML so that automated agents can interpret the page structure accurately. - **Layout Stability:** Layout stability prevents unexpected content shifts during automated scraping, reducing navigation errors and ensuring the agent maintains its position while processing information. - **AI Discovery File Existence:** This check confirms the presence of a canonical AI discovery file, typically located at `/agents.md`, which serves as a manifest for business rules and inventory access. The audit scoring methodology differs fundamentally from standard PageSpeed evaluations. Whereas conventional scores prioritize rendering speed and visual stability for humans, this new metric prioritizes semantic clarity and discoverability for machines. Merchants should note that resolving JavaScript rendering issues is often the first step, as heavy client-side scripts can obfuscate the accessibility tree and cause instability. ## How do you implement the required /agents.md file in Shopify? Passing the third criterion requires creating a custom `templates/agents.md.liquid` file within your Shopify theme to expose structured business data to AI agents at the `/agents.md` endpoint. In May 2026, Shopify quietly enabled native Liquid templates for this purpose to standardize agent interactions across the platform. To implement this, developers should navigate to Online Store > Themes > Edit code and add a new file named `templates/agents.md.liquid`. This file explicitly communicates your business structure, product availability, and purchasing policies to external AI systems. ``` # templates/agents.md.liquid name: {{ shop.name }} description: {{ shop.description }} version: 1.0 products: /api/products.json returns-policy: /policies/returns-policy contact: /contacts ``` By populating these Liquid variables, you ensure that when an AI agent scrapes your homepage at `domain.com/agents.md`, it immediately receives structured data regarding inventory and returns. This structured approach significantly improves the agent's ability to complete transactions autonomously. While some eCommerce frameworks utilize files like `/llms.txt` for large language model discovery, the PageSpeed Agentic Browsing audit specifically validates the `/agents.md` format for actionable agent guidance and policy enforcement. ## Why does Agentic Browsing matter for Shopify SEO and commerce? While the Agentic Browsing score is not currently a direct Google ranking factor, it acts as a strong proxy for technical SEO health and prepares your infrastructure for the Universal Commerce Protocol. Sites optimized for AI agents inherently possess better accessibility and cleaner DOM structures, both of which remain critical for traditional crawler indexing and core web vitals. Furthermore, implementing the `/agents.md` endpoint aligns your store with the Universal Commerce Protocol, a standard that allows agents to seamlessly interact with Shopify storefronts to process payments and manage carts. According to engineering analyses, enabling these endpoints positions Shopify stores to adopt default configurations that override legacy limitations, facilitating smoother integration with emerging commerce technologies. ## How can you test your store against this new audit today? You can verify your store's status by entering your homepage URL into the PageSpeed Insights tool, running the mobile audit, and reviewing the experimental Agentic Browsing section under the latest updates. Upon running the report, locate the Agentic Browsing card to view your current pass ratio. If you identify a low pass ratio, focus first on resolving any identified JavaScript rendering issues or missing semantic tags before implementing your custom Liquid template. Regular monitoring of this audit will help ensure your store remains compatible with evolving autonomous commerce ecosystems. ## References 1. [Lighthouse agentic browsing scoring - Chrome for Developers](https://developer.chrome.com/docs/lighthouse/agentic-browsing/scoring) 2. [PageSpeed Insights Update 2026: Agentic Browsing Explained](https://www.echovyn.com/blog/pagespeed-insights-agentic-browsing-update) 3. [Shopify Quietly Auto-Shipped /llms.txt, /agents.md and /.well-known](https://no7software.co.uk/blog/shopify-llms-txt-agents-md-ucp-defaults-engineering-override) 4. [Guide to llms.txt and agents.md for eCommerce - Shero Commerce](https://sherocommerce.com/blogs/insights/llms-txt-and-agents-md-for-ecommerce)