Quick Start

Get a score and the 8 fix files in a few minutes.

Prerequisites

• Node.js 20 or later
• A public website URL

Step 1: Scan

npx aeorank-cli@0.1.1 scan https://your-site.com

The published CLI crawls up to 50 pages by default, scores 12 dimensions, and writes 8 files. npm version 0.1.1 is canonical (-V may still print 0.0.1).

Step 2: Read the score

You’ll see a 0–100 score, a letter grade, and the 12 dimensions (high / medium / low). See 12 dimensions.

Illustrative shape (not a live transcript of your site):

Scanning https://your-site.com...
AEO Score: 42/100 (D)
12 dimensions · 8 generated files
→ 8 files written to ./aeorank-output/

Step 3: The 8 files

./aeorank-output/ contains exactly:

File	What it does
llms.txt	Agent-readable site map — not a citation guarantee
llms-full.txt	Longer extract
CLAUDE.md	Context for coding agents
schema.json	JSON-LD
robots-patch.txt	Allow rules for GPTBot and peers
faq-blocks.html	FAQ / speakable markup
citation-anchors.html	Heading anchors
sitemap-ai.xml	Sitemap stub

Not written: ai.txt, answers.json, report.html.

Step 4: Deploy

1. llms.txt / llms-full.txt / sitemap-ai.xml → site root
2. schema.json → JSON-LD in the document
3. robots-patch.txt → merge into robots.txt
4. HTML snippets → relevant pages

Rescan after deploy. Score change is technical readiness, not a citation timeline.

CI

To fail a PR when GPTBot is blocked, use the GitHub Action fail-on-crawler-block input. That flag is not on aeorank-cli@0.1.1.