Check whether your new site is built to be cited by AI assistants with an eight-point page audit
A free in-browser checklist turns published studies on AI citations into eight page checks a new site can fix, such as server-rendered text, structured data and a visible update date.
Evidence: The source shows its work: steps, screenshots, code or data you can inspect.
The business problem
Small new sites rarely rank on Google yet may still be cited by AI assistants. The author wanted to know what is within a new site's control instead of waiting for authority to build.
What was tried
The author first read several published studies on whether AI search engines cite the same pages that rank in Google, and found they often do not: a large share of AI Overview sources were outside the top 10 results, and different assistants cited largely different sources. From that they built a browser tool where you paste saved page HTML and it runs eight checks: body text is server-rendered rather than JavaScript-only; JSON-LD or microdata is present; content uses extractable lists; numbers name their sources; the page sits in a topic cluster with internal links; the canonical tag is an absolute URL; a visible last-updated date appears; and robots.txt does not block GPTBot, ClaudeBot, PerplexityBot or Google-Extended. It gives no single score and deliberately ignores llms.txt, since one cited log study found almost no requests for those files. The author ran it on their own pages, which scored 4 of 7 at the time, and fixed missing absolute canonicals, missing dates and a table that was empty in the served HTML.
What was reported (not reported)
The post does not report any traffic or citation gain. It says the author's own pages scored 4 out of 7 before fixes and that the problems found were corrected; whether AI assistants then cited the pages is not reported.
Limitations
The post is a synthesis of other people's studies, and it notes the studies disagree depending on whether citations are matched by URL, domain or brand. The checker's thresholds are the author's judgement, and the only real test subject was the author's own pages. Case data about time to first sale is self-reported and unaudited. The author sells a $19 founder toolkit and promotes their own tool, so treat it as partly promotional. It is not itself an AI workflow: it prepares pages for AI assistants to cite.
What you need
A published page whose HTML you can save, and access to its robots.txt and markup.
Sources
- DEV Community (20976093) ↗ Vendor case study, published October 10, 2026
Source published: October 10, 2026. Last reviewed here: October 17, 2026. Spot a mistake? Tell us.
Related workflows
Try to earn back an AI subscription with free Notion trackers and cheap digital products: $0 after 10 days and what to change
A founder-plus-AI-agents team sold free and cheap Notion trackers and a freelancer calculator with no ad budget, reached no sales in the sprint, and lists which distribution channels to drop.
Hand a founder's LinkedIn to a Claude Code agent pipeline for seven weeks, with a human approving every post, and read the median not the headline
A non-coder consultancy founder describes running LinkedIn through Claude Code with strategist, writer, editor and analyst roles, and reports a 25 times lift in total reach alongside a 1.3 times lift in the median post.
Install an open-source library of 200+ go-to-market skills for Claude Code covering SEO, outreach, ads research and competitor tracking
A maintained repository packages go-to-market tasks as skill files for coding agents, including SEO research, lead finding and enrichment, outreach drafting, content, competitor tracking and social monitoring.
Run four OpenClaw agents for content, leads and ops on Gemini's free tier, with a quality gate and a hard cap on requests
A solo developer in Taiwan describes four OpenClaw agents on a home machine that draft social posts, reply to comments, research, scan for leads and monitor endpoints, using Gemini's free tier at about 105 of 1,500 daily requests.