Not sure where to start? Take a guided tour.
AI Site Grader
AI Query Scanner
llms.txt GeneratorCrawler access
Can each crawler read your pages, or is it turned away?
- GPTBot
- OAI-SearchBot
- ChatGPT-User
- ClaudeBot
- anthropic-ai
- PerplexityBot
- Perplexity-User
- Bytespider
- Browser baseline
How does AI read your site?
We test requests using AI crawler user agents, inspect the responses and ask a model what it knows about you. Get findings, fixes and a technical grade, with the evidence for each.See what crawlers can read, what a model knows and what to fix.
Technical site grade
Your grade starts here.
No check has run yet. Your grade uses returned evidence.
WorksNeeds workObserved, not scored
Crawler access
How your site responds to AI crawlers.
Readable content
Text and metadata on pages checked.
Site files
Your robots.txt, llms.txt and sitemap.
Model knowledge
What a model knows about you, cold.
Off-site signals
Mentions, archives and sources elsewhere.
trakkr.ai/free-tools/ai-site-grader/yourbrand.comYour report gets its own page to share
What fills each cell.
Measured checks and supporting observations in five areas. Coverage follows the pages and responses in your report.
Readable content
Text and metadata on the pages the Agent reads.
- Visible text
- Page title
- Meta description
- Canonical URL
- H1 heading
- Structured data types
Site files
The public files that help crawlers find and read your site.
- robots.txt response
- Blanket robots rules
- Sitemap response
- Sitemap URL sample
- llms.txt
- Known public URLs
Model knowledge
What a model recalls about you without using search.
- Brand recall
- Products and services
- Who you serve
- Well-known facts
- Reputation signals
- Recall gaps
Off-site signals
Mentions, archives and public sources around your brand.
- Public mentions
- Returned source URLs
- Archived snapshots
- Hosting records
- Response headers
Google-Extended is a robots.txt policy token. Model knowledge and off-site signals are qualitative observations. Missing responses and outages are unavailable evidence. The report names every scored check and its source.
See a finished report.
Every report gets its own page you can share. Open one to see the findings and the evidence behind them.
Loading
stripe.com
Loading the public report and its evidence.
Reading the saved evidence
Open the report
Loading
linear.app
Loading the public report and its evidence.
Reading the saved evidence
Open the report
Loading
notion.so
Loading the public report and its evidence.
Reading the saved evidence
Open the report
Public reports, not customer case studies. A letter appears only when the saved evidence meets the grade method’s minimum coverage.
Questions,
answered.
The Agent reads your public pages, tests requests using AI crawler user agents, checks site files and asks a model what it recalls without search. Findings link to the underlying evidence. The technical grade covers only evaluable crawler access, readable content and site-file checks.
Ask anything else, like “Why would a crawler be blocked?”Ask anything else
The Agent answers from our docs, methods and pricing, and links to the source.
Then try these.
nike.com
“best running shoes”AI Rank Tracker
See where an OpenAI model names your brand for buyer questions, with each answer as proof.No web searchAEO Checker
Score technical readiness from homepage access, crawler rules and site files.Fixed 100-point rubricllms.txt Generator
Turn your public pages into an llms.txt you can copy or download.Copy or downloadChatGPT Ads Context Hints
Test buyer questions with an OpenAI model and build an ads planning pack.Planning packYour turn.
Type your domain. Get a report with the evidence and the fixes that matter.