The Complete 2026 Site Audit Checklist for AI Readiness

Over the past six months, we deployed 23 distinct technical upgrades across 52 websites. Not cosmetic changes — infrastructure-level implementations that make sites discoverable by AI agents, consumable by modern crawlers, and competitive in a search landscape that is shifting from keyword matching to content understanding.

This post is the complete checklist. Every item is something we have deployed, tested, and measured. Use it as your site audit framework for 2026.

Section 1: AI Agent Discovery

These files tell AI agents what your site does, what it covers, and how to consume your content. Without them, AI models have to infer this information from your pages — a process that is slow, incomplete, and error-prone.

[ ] /.well-known/agent.json — A2A Agent Card A JSON file declaring your site's name, description, capabilities, topics, content format, and available feeds. The A2A (Agent-to-Agent) protocol is Google's specification for inter-agent discovery. Deploy it even if you think AI agents are not visiting your site — they are, and this file helps them understand what they find.

[ ] /llms.txt — AI Summary File A plain-text Markdown file at your site root that provides a structured overview of your content. Include core topics, key URLs, content organization, and citation preferences. Think of it as a cover letter to every AI model that crawls your site.

[ ] /llms-full.txt — Full-Text AI File A comprehensive plain-text rendering of your most important content. This gives AI models with large context windows a single document to ingest instead of crawling dozens of pages. Cap it at 30,000-50,000 words to balance completeness with build time.

[ ] /humans.txt — Author Attribution A plain-text file identifying the people behind the site. Not AI-specific, but AI models parse it for author attribution and credibility signals. Include names, roles, and relevant credentials.

Section 2: Structured Data

Schema markup tells search engines and AI models what your content is, not just what it says. Every page on your site should have at least one schema type.

[ ] Article schema on all blog posts JSON-LD with @type: Article including headline, author, datePublished, dateModified, publisher, and image. This is the baseline for any content site.

[ ] FAQ schema on question-based content Any page that answers questions should have FAQPage schema with each question and answer marked up. This enables FAQ rich results and provides structured Q&A data for AI models.

[ ] ItemList schema on ranking/list pages Pages with ordered lists (rankings, comparisons, multi-part guides) should use ItemList schema. This enables carousel rich results and claims multiple SERP positions from a single page.

[ ] Dataset schema on original research Pages with original data — multi-variable analyses, survey results, calculated models — should use Dataset schema. This gets your data into Google Dataset Search alongside institutional sources.

[ ] DiscussionForumPosting schema on community content If your site has comments, forums, or discussion sections, DiscussionForumPosting schema can qualify your content for Google's Discussions carousel — a SERP feature previously reserved for Reddit and Quora.

[ ] SpeakableSpecification on key content Mark up the most important sentences on your page with SpeakableSpecification schema. This tells voice assistants and AI models which parts of your content to read aloud or quote.

Section 3: Syndication and Feeds

Feeds are how AI crawlers and aggregators consume your content at scale. Offer multiple formats.

[ ] RSS 2.0 feed (/feed.xml) The legacy standard. Still consumed by most feed readers and some AI crawlers. Include full content or substantial excerpts, not just titles.

[ ] JSON Feed 1.1 (/feed.json) The modern alternative. AI crawlers increasingly prefer JSON feeds because JSON parsing is simpler than XML parsing. Include post metadata, tags, and content.

[ ] Atom feed (/atom.xml) Some aggregators and crawlers prefer Atom over RSS. Low-cost to generate if you already have RSS.

[ ] Feed autodiscovery in HTML <head> Link tags advertising your feeds:

<link rel="alternate" type="application/rss+xml" href="/feed.xml" />
<link rel="alternate" type="application/feed+json" href="/feed.json" />

[ ] OPML blogroll (/blogroll.opml) An OPML file listing feeds you recommend. This participates in the blogroll network — a web of mutual feed references that drives discovery and establishes topical authority.

[ ] WebSub/PubSubHubbub hub registration Register your feeds with a WebSub hub (like Google's hub or a self-hosted hub) so that subscribers receive real-time push notifications when you publish new content.

Section 4: Performance APIs

These browser-native APIs improve user experience with zero JavaScript framework overhead.

[ ] View Transitions API Add @view-transition { navigation: auto; } to your global stylesheet. This gives your static site smooth, animated page transitions instead of jarring full-page reloads. Progressive enhancement — ignored by unsupported browsers.

[ ] Speculation Rules API Add a <script type="speculationrules"> block to your base template. This tells the browser to prerender pages before the user clicks, delivering instant navigation. Zero JavaScript libraries required.

[ ] Preconnect hints for third-party origins If your site loads resources from external domains (analytics, fonts, CDN), add <link rel="preconnect"> tags for those origins. This eliminates DNS and TLS latency for third-party requests.

[ ] Resource hints for critical assets Use <link rel="preload"> for above-the-fold images and critical CSS. Use <link rel="modulepreload"> for critical JavaScript modules.

Section 5: Image Optimization

Images account for the majority of page weight on most content sites. Proper optimization compounds across every page.

[ ] AVIF format with WebP fallback AVIF provides 30-50% better compression than WebP at the same quality. Use <picture> elements with AVIF as the primary source and WebP as the fallback. Browser support for AVIF exceeds 93% in 2026.

[ ] IPTC/XMP metadata on all images Embed creator, source, copyright, and description metadata directly in image files using ExifTool. This metadata travels with the image and provides attribution data to Google Images, stock platforms, and AI models.

[ ] Responsive images with srcset Serve appropriately sized images based on viewport width. A mobile user should not download a 2400px image when they need 600px.

[ ] Lazy loading with loading="lazy" Native browser lazy loading for below-the-fold images. No JavaScript required.

Section 6: Security and Trust

Security infrastructure signals professionalism to users, crawlers, and AI models.

[ ] /.well-known/security.txt — RFC 9116 A plain-text file with security contact information, expiry date, and disclosure policy. Required by an increasing number of compliance frameworks.

[ ] Security headers Content-Security-Policy, X-Content-Type-Options, X-Frame-Options, Referrer-Policy, and Permissions-Policy. These headers protect users and improve security audit scores.

[ ] HTTPS everywhere with HSTS All traffic over HTTPS with Strict-Transport-Security headers. This should already be done — if it is not, it is the single highest-priority item on this list.

Section 7: Alternative Discovery Channels

These are non-Google discovery channels that most sites ignore entirely.

[ ] Bluesky account with custom feed Build a custom algorithmic feed on Bluesky that curates content in your niche. Subscribers see your posts alongside relevant content from other accounts.

[ ] Brave Search Goggles Create a custom Goggle (ranking rule file) that boosts your content for relevant queries on Brave Search. Brave has 70+ million users.

[ ] Apple Shortcuts decision trees Build native decision-tree tools that run on 1.5 billion Apple devices and link back to your content. Zero hosting cost, zero app store approval.

[ ] Google NotebookLM source optimization Structure your content to be selected as a source by Google's NotebookLM. This means clear section headings, explicit data citations, and comprehensive topic coverage.

[ ] IndexNow protocol Notify search engines of content changes in real time via the IndexNow API. Bing, Yandex, and other search engines support it. Submit your IndexNow key and ping the endpoint on every publish.

How to Use This Checklist

Do not try to implement everything at once. Prioritize:

Week 1: AI Discovery. Deploy agent.json, llms.txt, and llms-full.txt. These have the highest impact-to-effort ratio.

Week 2: Structured Data. Add Article and FAQ schema to all content pages. Add ItemList to ranking pages.

Week 3: Performance. Add View Transitions and Speculation Rules. These are one-line deployments.

Week 4: Syndication. Add JSON Feed alongside existing RSS. Register with a WebSub hub.

Ongoing: Alternative channels. Build Bluesky feeds, Apple Shortcuts, and Brave Goggles as time permits.

Every item on this list was deployed across 52 websites and measured for impact. None are theoretical. All are practical.

The Resale Trap uses every item on this checklist as part of its technical infrastructure — powering the data-driven analysis that has been cited by AI models, researchers, and journalists. The 395-page analysis is available on Amazon. For the complete implementation guides, see The $100 Network and The $20 Dollar Agency.