What this tool checks
Audit article AEO and SEO reviews one public article URL and returns a prioritized report across three layers:
- System checks related to article discovery, crawling, and machine readability, such as title, meta description, H1, canonical, noindex, image alt coverage, internal links, robots.txt, sitemap, and llms.txt.
- Content SEO checks such as search intent fit, topic coverage, freshness, evidence, author trust, and readability.
- AEO, AIO, and LLMO checks such as extractable answer blocks, FAQ structure, comparison tables, cited statistics, author attribution, llms.txt, and AI crawler access.
How to use it
Paste the article URL and run the audit. A target keyword is optional. When you add a keyword, the tool also checks the current SERP and compares up to three top organic pages.
Use the report as an editing checklist: fix blocking crawl/index issues first, then improve the article structure, evidence, FAQ coverage, and AI-search readability.
Output
- Overall score and top fixes.
- Checks with pass, warn, fail, or unknown status.
- Technical SEO, Content SEO, and AEO / AI SEO check groups.
- Optional SERP comparison when a target keyword is supplied.
- Source URLs and caveats.
Check Items
Checks are split between system-determined checks and LLM structured judgements. User-facing labels and recommendations are generated by the UI, not stored in the run data.
| Area | Check | Method | What it looks at |
|---|---|---|---|
| Technical SEO | HTTP status | System | Whether the article URL returns a stable 200 response |
| Technical SEO | Canonical URL | System | Whether canonical exists and matches the fetched URL intent |
| Technical SEO | Robots meta | System | Whether noindex or similar directives block discovery |
| Technical SEO | JSON-LD structured data | System | Whether JSON-LD is visible in fetched HTML |
| International SEO | Hreflang integrity | System | Hreflang codes, self-reference, x-default, and canonical alignment |
| On-page SEO | Title length | System | Whether title length fits the target market range |
| On-page SEO | Meta description | System | Whether meta description length fits the target market range |
| On-page SEO | Image alt coverage | System | Whether informative images have alt text |
| On-page SEO | Internal links | System | Whether the article links to related pages |
| On-page SEO | Cited evidence and statistics | System | Sourced statistics, expert quotes, and external links |
| On-page SEO | Author and freshness signals | System | Author signals, published date, and updated date |
| Content structure | H1 count | System | Whether exactly one H1 is present |
| Content structure | Heading coverage | System | Whether H2/H3 sections divide the topic clearly |
| Content structure | Heading hierarchy | System | Whether heading levels skip in an unusual way |
| Content structure | AEO content pattern coverage | System | Definition, steps, FAQ, comparison table, pros/cons, or list formats |
| Content structure | Short answers after headings | System | Whether question-style headings are followed by concise answers |
| Content structure | Unnatural writing pattern risk | System | Dash punctuation, template phrases, abstract terms, and weak intensifiers |
| Content structure | Japanese editorial naturalness | System | Japanese abstract wording, formulaic structure, and mechanical leftovers |
| AI crawler access | Search and AI crawler access | System | Whether Googlebot, Bingbot, GPTBot, PerplexityBot, ClaudeBot, and related crawlers are allowed |
| For AI crawlers | llms.txt | System | Whether /llms.txt is served as text or Markdown |
| For AI crawlers | AI-useful structured data types | System | Article, BlogPosting, FAQPage, HowTo, and related schema types |
| AI SEO | Google AI Overview readiness | System | Structured data, short answers, evidence, authorship, and internal links |
| AI SEO | ChatGPT citation readiness | System | Answer format, freshness metadata, cited statistics, and OpenAI crawler access |
| AI SEO | Perplexity citation readiness | System | FAQ/HowTo, self-contained answers, citations, schema, and PerplexityBot access |
| AI SEO | Claude and Copilot readiness | System | Factual density, definitions, freshness metadata, Claude access, and Bingbot access |
| Content quality | Search intent fit | LLM structured judgement | Whether the article answers what the target query expects |
| Content quality | Topic coverage | LLM structured judgement | Whether important subtopics are missing |
| Content quality | Specificity and usefulness | LLM structured judgement | Steps, examples, numbers, and decision criteria |
| Content quality | Evidence quality | LLM structured judgement | Whether important claims are backed by verifiable evidence |
| Competitive analysis | Competitive content gap | LLM structured judgement | Gaps versus SERP competitors when competitor data is available |
How the overall score is calculated
The overall score is not assigned by the LLM alone. The final score combines checks with an LLM content-quality score: 65% checks and 35% content quality.
| Component | Weight | What it covers |
|---|---|---|
| Checks | 65% | HTTP status, noindex, canonical, H1, internal links, robots.txt, sitemap, llms.txt, structured data, AEO content patterns, and AI crawler access |
| Content quality | 35% | Search intent fit, topic coverage, specificity, evidence quality, readability, originality, and extractable explanations for AI search |
Checks are scored with a weighted average of status and impact. Status values are pass = 1.0, warn = 0.6, unknown = 0.45, and fail = 0. Impact weights are high = 3, medium = 2, and low = 1.
The final score is capped when important blockers fail. Crawl or index blockers such as HTTP status, noindex, robots.txt, or AI crawler access cap the score at 49. High-impact article structure or international SEO failures such as H1, canonical, or hreflang cap the score at 69. This means a strong article cannot receive a high score if search engines or AI crawlers cannot reliably read it.
Limitations
This is not a Search Console, analytics, PageSpeed, rank-tracking, or AI Overview measurement tool. It uses public fetches, SERP data when requested, and AI synthesis. If no JSON-LD is found in fetched HTML, the report marks schema as unknown because rendered JavaScript may inject structured data.