We open-source our agentic-commerce tooling.
Anyone can claim “agent-ready.” We publish the tools that prove it — MIT-licensed, honest limitations in every README, benchmarks you can reproduce. Starting with agentready: scan any store and see exactly what AI shopping agents see.
No email needed for the score · ~15 polite requests · MIT
$ npx @mntglobal/agentready aeropress.comB+ 88/100 — 7 fixes to become fully agent-ready████████████ Agent Access 15/15███████████░ Structured Data 23.5/25██████████░░ Protocol Endpoints 11/13██████░░░░░░ Data Freshness 4/8✓ PE-02 UCP manifest — discoverable by Google'sagentic surfaces (v2026-04-08)⚠ SD-04 Identifiers missing: gtin/mpn
29 checks. 100 points. One honest grade.
agentready fetches your store the way an agent does — no JavaScript, no mercy — and grades what it finds, with evidence quoted from your actual pages.
Structured data depth
Product/Offer/ProductGroup JSON-LD — price, availability, identifiers, ratings.
Agent access policy
robots.txt scored by purpose: blocking a training bot is fine; blocking a shopping agent costs sales.
Protocol discovery
ACP feed-readiness, Google UCP manifest, live MCP endpoints — the 2026 transaction layer.
Feeds, freshness & AEO
Catalog endpoints, sitemap freshness, price consistency, citability, a11y quick pass.
How the 100 points break down.
Weights reflect what actually moves agent visibility in 2026 — protocol discovery weighs more than accessibility because that's where stores lose sales today.
Structured Data
Product/Offer/ProductGroup JSON-LD: price, availability, GTIN, ratings
Protocol Endpoints
ACP feed-readiness · Google UCP manifest · live MCP endpoints
Agent Access
robots.txt by purpose-class, llms.txt, no bot-walls
Product Feeds
Sitemap product coverage, open catalog endpoints, feed signals
Machine Readability
Content in the initial HTML — agents don't run your JavaScript
AEO Citability
Metadata, FAQ markup, shipping/returns policy discoverability
Data Freshness
Sitemap lastmod, HTTP validators, schema-vs-page price consistency
Accessibility
Quick pass: alt coverage, lang/labels, heading order
Every check ships with evidence quoted from your pages and a concrete fix — see the full 29-check reference
Proof over promises. Honesty over hype.
The agentic-commerce protocols are months old. Everyone claims expertise — we'd rather show working code.
Inspectable expertise
Our protocol research, spec pins and probe data ship in the repo. You can audit what we scan — closed-source competitors can't say that.
Honest by design
Every README has a limitations section. Checks that can't run are excluded from your score — never counted against you. If no vendor consumes llms.txt, the report says so.
Free tools, expert fixes
The scanners are MIT-licensed forever. Our business is the engineering behind the fixes — platform builds, protocol integrations, embedded agents.
agentready is the first of four.
Every tool maps to a real problem we fix for clients — no orphan repos, no abandonware.
vibecheck
Production-readiness auditor for AI-built stores — injection points, exposed secrets, price-tamper logic, load behaviour. Born from our AI Cleanup Lab.
retail-mcp
Open-source Retail MCP server kit — expose catalog, inventory and pricing to AI agents securely. Shopify adapter first.
acp-testkit
Agentic-checkout conformance suite + a mock shopping agent — test your ACP/UCP integration before a real agent embarrasses you.
Follow along: github.com/MnT-Global · or get release notes by email
Use it your way.
Same open-source engine everywhere — pick whichever fits how you work.
In your terminal
One command — npm downloads and runs it. Your machine fetches your store directly; nothing touches our servers.
npx @mntglobal/agentready your-store.comCLI docs on GitHub In your browser
No terminal needed — the same engine runs on our site. Free score, no email required.
mntfuture.com/agentreadyAs source code
Clone it, read every check, run the 152 tests, or build your own tooling on the core engine (MIT).
git clone https://github.com/MnT-Global/agentreadyBrowse the repository Fair questions, honest answers.
Yes — MIT-licensed, forever. The CLI is free, the hosted scan is free, and no email is needed for the score. Our business is the engineering behind the fixes, not the scanner.
With the CLI: nothing reaches us — it runs on your machine and fetches your store directly. With the web scanner: our server fetches your public pages to compute the score shown to you; we only keep your email and grade if you request the fix plan.
Anyone can propose a change (fork → pull request), but nothing merges without our review, passing tests, and a signed-off commit. Only the MnT team can push to the main branch.
The agentic-commerce protocols are months old and full of bold claims. Publishing working, inspectable tools is how we prove we're practitioners — and stores that discover gaps often ask us to fix them. Honest exchange.
agentready finds the gaps. We build the fixes.
Run the free scan, then bring the report to a free architecture workshop — a senior engineer maps every finding to a concrete fix plan for your platform.