inemaSEO 4 — In the AIs: AEO, GEO, and sites used by agents|INEMA.CLUB-PRO

Lesson 6 · 17 min

AI readiness rubric and its caveats

An AI readiness score helps you find what's broken, as long as you know how it's calculated. No score is official: each scanner picks its own criteria and weights. Here you compare two open INEMA rubrics, read what weighs most, and learn what the score does not prove.

Start the lesson
Inspector with a clipboard checks the entrance of a commercial building, with a ramp and a glass door.

In this lesson you will

  1. 1How WebMCP Readiness and agente-ready calculate the score
  2. 2The criteria that weigh most: bot access, content, entity
  3. 3The caveats: WebMCP weight, questions and answers, small sample
  4. 4Fix what weighs most and run it again

Goal: run an AI readiness score, read the criteria that weigh most, and know what the score does not prove.

01How a score is born

Click each block.

Click each block in the diagram to see what it does.

The score helps you prioritize fixes. Citations are measured in lessons 2 and 3.

02The two rubrics and the caveats

Click each tab.

1. WebMCP Readiness

The INEMA scanner adds up four blocks: WebMCP 30%, SEO 30%, GEO 20%, and AEO 20%. Each finding is worth pass, warning, or fail; an informational-only finding is worth 0.7.

GEO looks at entity, substantial content (250+ words), authorship and date, sources, and semantic HTML. AEO looks at questions in headings, direct answers, lists, and tables.

03Run it and read the score

agente-ready only reads public pages, with no API (detailed in course 7). WebMCP Readiness (webmcp.inema.pro) opens the page in a disposable browser and goes deeper only on the URL you provide.

agente-ready (inemaseo-sistemas, MIT)
git clone https://github.com/inematds/inemaseo-sistemas && cd inemaseo-sistemas/ferramentas/agente-ready
python3 agente-ready.py https://seusite.com --saida saida/seusite
# output: relatorio.html, agente-ready.json and criterios.csv
# the weights live in pesos.json: disagree, change them and run again
notas.md · log
# AI readiness — <seusite.com>
| date | tool and version | score | 3 heaviest failures | fixed |
|---|---|---|---|---|
| dd/mm/yyyy | agente-ready 1.0.0 | <n> | <failure 1 (weight)>, <failure 2>, <failure 3> | <which> |
Example from the repository

agente-ready 1.0.0 on INEMA sites, home + 4 pages:

site                         nota  entidade  conteudo  descoberta  webmcp
inema.club                   93    86        81        100         100
news.inema.pro               72    0         71        100         29
eventos.inema.pro            69    0         62        100         29
guia REA (github.io)         54    0         62        17          29

04Practice: your site's score

It can be your own site or a client's who gave permission.

E1 · Score and 3 failures

20 min

Run agente-ready (or WebMCP Readiness) on the site and write down the score, the tool with its version, and the 3 heaviest failures.

Done when: notas.md has the row with date, tool and version, score, and the 3 failures with the weight of each.

E2 · One measured fix

40 min

Fix the heaviest failure within your reach (for example, the entity on the home page), publish, and run it again.

Done when: notas.md has the second row with the new score and the fixed criterion marked as ok.

05Test what you learned

The explanation appears after your answer.

Question 1 of 4

The site has flawless SEO and a low overall score in WebMCP Readiness. What is the likely explanation?skill

06Takeaway

A score is a rubricEach scanner picks its own criteria and weights; prefer public weights.
Prioritize by weightBot access, content in the HTML, and entity come before details.
2026 yardstickA high score is not a citation. WebMCP and questions in headings are preparation and clarity, not a ranking factor.

Go further

WebMCP Readiness (INEMA scanner)inemaseo-sistemas: agente-ready and pesos.json (MIT)Google: create helpful content (English)
Previous lessonNext lesson