AEOrank scores crawler access and extractability 0–100 across 12 dimensions (aeorank-cli@0.1.1), then generates the 8 files that CLI actually writes. Crawler allowlists matter; the GitHub Action can fail a PR if GPTBot is blocked. llms.txt is agent-docs hygiene, not a citation guarantee.
Others monitor. AEOrank fixes.
Star the repo if this is useful — 12 dimensions, 8 files, aeorank-cli@0.1.1.
Website · Docs · CLI scan · Dashboard (experimental) · GitHub Action · GitHub App
Star → try CLI → optional Audit
Star this repo if you want a deterministic CI gate and 8 fix files — not another closed monitor.
npx aeorank-cli@0.1.1 scan https://your-site.comNeed it delivered? Mindtrades offers an AI Visibility Audit (separate paid service). AEOrank stays MIT/free: aeorank.dev.
30-second scan · 12 dimensions · 8 files · Action can fail the PR if GPTBot is blocked
AI search engines now drive 40% of web discovery. ChatGPT converts visitors at 15.9% — higher than Google organic. But traditional SEO tools don't check what AI engines actually look for.
AEOrank does.
Profound, Peec, Otterly, Scrunch, and Athena are closed monitoring SaaS (prompt tracking, brand mentions), typically $89–$499/mo. AEOrank is MIT, CLI/CI-first, and writes the 8 fix files. We do not promise citations.
| AEOrank | Profound / Peec / Otterly / Scrunch / Athena | |
|---|---|---|
| Price | CLI free / Pro $29 | typically $89–$499/mo |
| Open source | ✅ MIT | ❌ closed |
| CLI + CI | ✅ Action + App | ❌ dashboards |
| Framework plugins | 11 | 0 |
| Generates AI files | ✅ 8 files | ❌ mention reports |
| Scoring | 12 dimensions (aeorank-cli@0.1.1) | LLM prompt tracking |
| Citation guarantee | ❌ none | ❌ they monitor after the fact |
| Self-hostable | ✅ | ❌ |
npx aeorank-cli@0.1.1 scan https://your-site.comJSON output for scripting:
npx aeorank-cli@0.1.1 scan https://your-site.com --format json | jq '.score'Install the AEOrank GitHub App on your repo. Every PR automatically gets an AEO score as a Check Run — no YAML, no config. The Check fails if GPTBot (or ClaudeBot, PerplexityBot, Google-Extended) is blocked. fail-below (score threshold) is not available on the App yet — use the Action for that.
name: AEO Score
on: [push, pull_request]
permissions:
checks: write
pull-requests: write
contents: read
jobs:
aeo:
runs-on: ubuntu-latest
steps:
- uses: vinpatel/aeorank-action@v1
with:
url: https://your-site.com
fail-below: 50
fail-on-crawler-block: true # fail the PR if GPTBot is blockedPublished CLI weights are high / medium / low — not a 36-row percentage catalog.
| Weight | Dimensions |
|---|---|
| high | llms.txt, schema.org markup, content structure |
| medium | AI crawler access, answer-first, FAQ & speakable, E-E-A-T, meta descriptions, citation anchors |
| low | sitemap, HTTPS & redirects, page freshness |
ai.txt is not scored and not generated. A 36-row catalog exists in repo source and is archived in docs until npm CLI dimensions.length matches it.
These are the files npx aeorank-cli@latest (npm 0.1.1) actually writes (generateFiles() — no stubs, no ai.txt):
| File | Purpose |
|---|---|
llms.txt |
Site summary for LLM crawlers (llmstxt.org spec). Agent-docs hygiene, not a citation guarantee. |
llms-full.txt |
Full-context version with Q&A pairs and entity disambiguation |
CLAUDE.md |
Markdown context file for Claude |
schema.json |
JSON-LD structured data |
robots-patch.txt |
AI-specific robots.txt rules (GPTBot, ClaudeBot, PerplexityBot, Google-Extended) |
faq-blocks.html |
FAQ with schema.org + speakable markup |
citation-anchors.html |
Deep-linkable citation anchors |
sitemap-ai.xml |
AI-optimized sitemap |
The CLI JSON includes a dimensions array of length 12. GitHub has a v1.0.0 tag that is not on npm; do not treat that tag as the published CLI. Binary -V may still print 0.0.1 — npm 0.1.1 is canonical.
That gate is the GitHub Action input fail-on-crawler-block (not a flag on aeorank-cli@0.1.1). It fails the Check when GPTBot, ClaudeBot, PerplexityBot, or Google-Extended is disallowed in robots.txt. A missing robots.txt is unknown, not blocked.
Drop-in AEO file generation. One config, the same 8 files the CLI writes, zero maintenance.
npm install @aeorank/next # or @aeorank/astro, @aeorank/nuxt, etc.// next.config.ts
import { withAeorank } from "@aeorank/next";
export default withAeorank({
siteName: "My Site",
siteUrl: "https://example.com",
description: "What my site does.",
});
// → The 8 generated AEO files now served at your site rootThe working score path is the CLI above — no account. The hosted dashboard at app.aeorank.dev is experimental (Clerk Development), not a Production signup flow:
- Add a public URL → 12-dimension breakdown (when the hosted instance is available)
- Download the 8 generated files as a ZIP
- Free tier: 1 site, 3 scans/month (CLI stays unlimited)
- Pro $29: 5 sites, 50 scans/month
| Package | Description |
|---|---|
aeorank-cli |
CLI — npx aeorank-cli@0.1.1 scan <url> |
@aeorank/core |
Core scanning + scoring engine |
@aeorank/next |
Next.js plugin |
@aeorank/astro |
Astro integration |
@aeorank/nuxt |
Nuxt module |
@aeorank/remix |
Remix plugin |
@aeorank/sveltekit |
SvelteKit plugin |
@aeorank/gatsby |
Gatsby plugin |
@aeorank/shopify |
Shopify Hydrogen plugin |
@aeorank/11ty |
Eleventy plugin |
@aeorank/vitepress |
VitePress plugin |
@aeorank/docusaurus |
Docusaurus plugin |
See DEMO.md for today's auto-generated scan.
Last updated: September 20, 2026
Star if you want a deterministic CI gate and the 8 fix files aeorank-cli@0.1.1 actually writes — not another mention dashboard.
We ship fast and merge fast. New framework plugins, new scoring criteria, and real-world bug reports are especially welcome.
- 👀 Browse good first issues — concrete, small, well-scoped.
- 📖 Read CONTRIBUTING.md — repo layout, PR expectations, local setup.
- 💬 Ask in Discussions — ideas, questions, show-and-tell.
- 🔒 Report vulnerabilities privately via Security Advisories.
git clone https://github.com/vinpatel/aeorank.git
cd aeorank
pnpm install
pnpm test # core + CLI + pluginsIf AEOrank saved you a closed monitoring subscription, the best thank-you is a star plus sending it to one person who also needs a fix/CLI/CI tool.
MIT — built by Vin Patel, sponsored by Linx Agency & Mindtrades.