TrizCom has published 417 posts. In the markup an answer engine actually reads, not one of them was written by a person.
Twelve signals, weighted toward attribution and identity rather than crawl mechanics, computed from raw HTML fetched live. Nothing on this page is typed by hand; every number below is written by the analyzer into a JSON file that this page renders.
Worth saying first, because it changes what the work is. This is not a rebuild. The foundations are in place and the gaps are narrow and specific.
Each one is checkable in your own view-source in under a minute. I have quoted the exact values rather than describing them.
The five pages I sampled, scored independently.
A captured audit is unfalsifiable, so here is the same analyzer as a live endpoint. Point it at any public page, including a client's, including a competitor's. It fetches one page plus robots.txt, llms.txt and sitemap.xml, and scores what it finds.
The endpoint fetches a URL you supply, so it is hostile input by definition. It allows http and https only, resolves DNS inside the socket's own resolver so the address that was checked is the address that gets connected to, re-validates every redirect hop, and caps both body size and request time. The DNS rebinding hole that the naive version of this has is closed, and there is a test for it.
In the order that moves citability fastest, not the order that is easiest to bill.
Author objects that are people, with their own entity, tied to each post. This is the one change that turns 417 anonymous posts into a body of work by named experts.
One organisation entity, one canonical name, one https URL, one set of profiles, linked by @id so the publisher of a post and the agency on the homepage are the same thing.
Posts already titled as questions are answer-engine targets already. FAQ markup is the short path from that content into an answer box.
A baseline, then the same prompts re-run on a schedule, reported as movement a C-suite reader can act on. Not a screenshot of a chatbot.