Sample deliverable
AI-readiness scan: chatwith.tools
Public data only, run October 27, 2026. This is the technical section of an audit. I have no connection to chatwith.tools; I picked it as a public example.
Score: 8 out of 10
Good foundations: an llms.txt file, a sitemap, Organization, SoftwareApplication and AggregateOffer schema, and about 1,200 words of crawlable text on key pages.
Finding 1 (high impact): AI crawlers are blocked
robots.txt blocks GPTBot, ClaudeBot, Google-Extended and CCBot. These crawlers feed model training and some AI answer features, so blocking them can make ChatGPT, Claude and Gemini less likely to know about the product. OAI-SearchBot and PerplexityBot are allowed, so live search citations can still happen.
Fix: make the policy a deliberate choice. If the goal is to be recommended, allow GPTBot and ClaudeBot on marketing pages and keep app and user-data paths blocked.
Finding 2: no FAQ schema
Answer engines pull out question-and-answer blocks. Adding FAQPage schema to the pricing and comparison pages is a quick win.
How it compares
| Site | Score | What I found |
|---|---|---|
| chat-data.com | 9/10 | Best prepared of the four. |
| chatwith.tools | 8/10 | Has llms.txt, a sitemap and product schema, and about 1,200 words of crawlable text. But robots.txt blocks GPTBot, ClaudeBot, Google-Extended and CCBot. No FAQPage schema. |
| openassistantgpt.io | 6/10 | No schema markup and no llms.txt. |
| seobotai.com | 6/10 | No schema markup and no llms.txt. |
Public scan, October 27, 2026. Technical readiness only; a full audit adds the prompt testing.
What a full audit adds
- 40 prompts tested against named competitors, run on two different days.
- The sources each AI engine cites.
- A ranked 30-day fix plan and a recorded walkthrough.