Agentic SEO Check Review 2026: Free SEO and AEO Audits

Agentic SEO Check is a free, open-source audit tool from RIVI Software that checks a page or site for search engines and AI answer engines, and gives the exact edit that fixes every finding. It runs in the browser, as a CLI or as a hosted MCP server in Claude, Cursor or Codex (agentic-seo.rivi.software, checked 13 November 2026). Our desk verdict: the most useful free AEO audit we have seen for technical teams, with thin traction so far. It scores 4.1 out of 5.

For the quick facts, see the Agentic SEO Check listing on Digital Marketing Supermarket. Browse related AI SEO, AIO and GEO tools, or our AI SEO hub.

What Agentic SEO Check Is

RIVI Software is a small Dutch company whose main product is RadioTools, software for radio stations. Agentic SEO Check is its open-source side project: the GitHub repository was created on 8 November 2026 and the tool launched on Product Hunt on 12 November 2026, announced by @rivi_software.

Two audits run side by side, each with its own score: 184 SEO rules covering titles, canonicals, structured data, links, performance and site-wide checks, and 32 AEO rules covering whether AI crawlers may fetch a page and answer engines can quote it. The output is built for coding agents: each finding carries a rule ID, severity, CSS selector, the evidence as served and the edit that fixes it, with words only you should write left as placeholders.

Evidence: What Is Real and What Is Claimed

Verifiable: the code itself, MIT-licensed and public; the rule catalogue on the site (216 rules); the Product Hunt launch; a named maker with its own company site and X account. Because the rules are open, you can inspect exactly what is checked.

Thin: traction. The GitHub repository had 2 stars and 1 fork on 13 November 2026, and @rivi_software had 6 followers; the launch post drew 13 likes and 13 replies. Some directories quote older rule counts (about 160 SEO and 30 AEO); we used the live site’s 184 and 32. We ran the web audit page only as a demo, not a full site audit.

Who Agentic SEO Check Is For

• Technical SEOs who want a fast audit that names the fix, not just the problem.
• Developers and coding-agent users who want Claude Code, Cursor or Codex to audit, fix and re-audit a build folder.
• Content and growth marketers checking whether AI crawlers such as OAI-SearchBot and PerplexityBot are blocked.
• Agencies that need a free first-pass audit for prospects.

It is a poorer fit for marketers who want dashboards, rank tracking or AI-citation monitoring, or who cannot run Node.

Key Features

SEO audit. 184 rules, led by 49 site-wide checks, 17 performance, 14 structured data and 13 link rules, with hreflang, headings, images, Open Graph and JavaScript-rendering checks.

AEO audit. 32 rules: 9 on AI crawlers in robots.txt, 7 on snippet and answer controls, 7 on authors, dates and entities, 6 on readable pages and 3 on llms.txt. robots.txt is read for 25 crawlers, split into search index, AI answers and model training, so you can allow ChatGPT search while blocking GPTBot training, for example.

Three ways to run it. A web audit (a page, or a site from its homepage, up to 100 pages at 2 MB of HTML each); a hosted MCP server with no sign-in; and an npm CLI (Node 22.19 or later) with optional JavaScript rendering and Lighthouse, JSON, Markdown or PDF output and CI exit codes.

Audit, fix, check again. Re-running shows what is fixed, what is still open and what is new.

Pricing

Option	Price	What you get
Web audit	Free	A page or up to 100 pages; JSON, Markdown or PDF
Hosted MCP server	Free, no sign-in	Audit tools inside Claude, Cursor, Codex and other MCP clients
CLI / npm package	Free (MIT)	Local audits of URLs, build folders or raw HTML; optional rendering and Lighthouse

Source: agentic-seo.rivi.software and GitHub, checked 13 November 2026. No paid tiers or affiliate programme.

Caveats and Limitations

• Audit only. No rank tracking or AI-citation monitoring.
• No JavaScript in the web audit. Client-rendered sites need the CLI with rendering.
• Developer-first. The best features assume Node or an MCP-capable coding agent.
• Continuity. A brand-new side project with one small maintainer; support is via GitHub.
• Desk review only.

Our Desk Score

• Features: 4.0/5. 216 rules, AI-crawler matrix, exact fixes, MCP and CLI; no tracking.
• Evidence: 3.0/5. Named maker and inspectable open code; minimal traction and no independent reviews yet.
• Risk and fit: 3.0/5. Read-only and free, so low data risk; but developer-first for a marketing audience, and an early side project.
• Pricing: 5.0/5. Completely free and MIT-licensed.
• Ease: 3.5/5. The web audit is instant; the CLI needs Node 22.19+.
• Overall (v2): 4.1/5. Weighted average 3.625, plus the v2 scale adjustment (no gate binds).

This is a desk review based on the Agentic SEO Check site and rule catalogue, its GitHub repository, its Product Hunt page and the launch post on X, checked on 13 November 2026.

Frequently Asked Questions

Is Agentic SEO Check free?

Yes. The web audit, the hosted MCP server and the npm package are free, and the code is MIT-licensed on GitHub. No account or API key is needed. Checked 13 November 2026.

What does the AEO audit check?

32 rules on whether AI answer engines may fetch and quote a page: AI crawler rules in robots.txt (25 crawlers, split into search, AI answers and model training), snippet and answer controls, authors, dates and entities, readable pages and llms.txt.

Does Agentic SEO Check track AI visibility over time?

No. It audits pages as served and tells you what to fix. To measure how often AI assistants mention your brand, pair it with a visibility tracker.