A Dedica AI agent website check answers a question most businesses have never asked: can an AI agent actually use your website? Not read it, not summarise it — use it. Find the pricing, compare the plans, create an account and check out, the way a human visitor would. Dedica, launched on Product Hunt on 27 October 2026, runs 125 checks across discovery, access, usability and payments, scores your site out of 100, and hands you a fix for every place agents fail. The score itself is free, takes under a minute and needs no sign-up. Our verdict up front: this is the most practical agent-readiness tool we have seen for small and mid-sized teams, because it tests behaviour rather than publishing a checklist. But a score is not a strategy, and the fixes it suggests deserve a human decision before they ship.
Dedica AI Agent Website Check: What It Actually Tests
The idea behind Dedica came from a failed sign-up. Its maker watched an AI agent try to register a team for a software tool: the agent found the pricing page, got an empty page because the plans only loaded with JavaScript, then hit a bot check on sign-up and gave up. The agent did not report an error to anyone. It simply told its user it could not do that. That silent failure is the whole problem Dedica exists to expose. When an agent cannot complete a task, there is no angry support ticket, no abandoned-cart email, no second chance. The user just goes elsewhere, and you never learn you lost them.
Dedica tests your site the way agents like ChatGPT, Claude, and other browsing agents actually use it. You enter a domain, and within a minute you get a score out of 100 plus a ranked list of fixes, ordered by how much each one moves the score. A sample report on the official site shows 42 fixable issues, with the top items carrying estimated gains of four to six points each. That ranking matters: agent readiness is a long list of small defects, and teams need to know which three to fix first, not stare at forty.
The 125 Checks, in Four Buckets
Dedica groups its checks into four categories, and the split is a useful mental model even if you never run the tool:
- Discovery. Can agents find your site, understand what it offers, and reach the pages that matter? This covers machine-readable basics: whether key content is present in the HTML the server sends, whether your documentation is discoverable from the homepage, and whether structured data helps an agent make sense of your offering.
- Access. Can agents actually reach your pages, or does your security stack turn them away? Bot checks, challenge pages and aggressive rate limiting are the usual culprits. Dedica flags cases such as a sign-up page returning a 403 challenge page to agents.
- Usability. Can an agent complete the journey once it is in? The classic failure here is content that only exists after JavaScript runs: pricing tables, plan comparisons and feature lists that render in the browser but arrive empty to a headless agent.
- Payments. Can an agent finish a purchase or sign-up? This is where agent commerce is heading fastest, with checkout protocols from OpenAI, Google and Shopify now in play, and it is where a broken step costs real revenue rather than just visibility.
The concrete examples from Dedica’s own sample report show why this beats a generic audit. The top four fixes were a bot check blocking agents at sign-up, pricing that needs JavaScript to load, no published OpenAPI specification, and no MCP server. Each is a real, checkable defect with a named page and a named remedy, not a vague recommendation to “improve your content”. For a business owner, that specificity is the difference between a report that gets filed and a report that gets fixed.
Pricing: a Free Score and a $30 Fix Kit
The pricing is refreshingly simple, and we verified it against both the Product Hunt launch post and the official site. The scan and the score are free, with no sign-up required. The paid product is the fix kit: at launch pricing it costs $30 as a one-time payment per site (the regular price is $50). For each issue, the kit gives you one instruction file that says what to change and how to check the change worked, written so it can be handed to a developer or pasted directly into Claude Code, Codex, Cursor or GitHub Copilot.
That delivery format is a smart choice. Most audit tools produce PDFs that a developer must interpret; Dedica produces work orders a coding agent can execute. The economics are equally plain: $30 is less than one hour of any developer’s time, so the kit pays for itself if it saves a single round of “what does this finding actually mean?” There is also a public API and an MCP server, so your own agents can check sites themselves rather than clicking through a dashboard.
One caution on price: launch pricing is, by definition, temporary. The $50 regular price is the number to budget against if you are reading this after the launch window. And the kit is priced per site, so an agency auditing twenty client sites is looking at twenty kits, not one — worth knowing before you promise a client a fixed-price audit.
Who Dedica Is For — and Who Should Skip It
Dedica is built for businesses whose websites do real work: SaaS sign-ups, service bookings, e-commerce checkouts, lead capture. If your site’s job is to convert a visitor into a trial, a booking or a sale, then an agent that cannot complete that journey is a leak in your funnel, and this tool finds the leak. It is also a natural fit for agencies: a free score is an excellent first deliverable in a client conversation, and the fix kit gives the engagement a concrete scope.
Who should skip it? If your site is a simple brochure with a phone number as its only conversion path, agent readiness is not your bottleneck yet. Likewise, if you have not yet fixed the basics — slow pages, broken mobile layouts, missing analytics — fix those first. Agent traffic is growing, but human traffic still pays the bills, and a site that frustrates people will frustrate agents too. Dedica is a second-layer audit, not a substitute for the first layer.
Watch-Outs and Honest Limitations
Three limitations deserve honest airtime. First, Dedica’s headline numbers are marketing figures, not independently verified research. The official site claims only 6% of business websites are ready for AI agents and that ready businesses are recommended 1.9 times more often by AI agents. Treat those as directional, not gospel: the 6% figure comes from Dedica’s own testing, and the recommendation claim depends heavily on which agents, which queries and which businesses were measured. The tool’s value does not depend on those numbers being exact, but your business case should not either.
Second, a score is a snapshot of one domain at one moment, not a monitoring service. Sites change constantly — a new bot rule, a redesigned pricing page, a checkout provider update — and each change can reintroduce a failure. The official workflow anticipates this (“scan, fix, scan again”), but there is no indication of scheduled re-scans or alerting in the launch offering. Treat the first scan as a baseline and re-run it after every significant release.
Third, and most important: not every fix should be applied blindly. Dedica’s top recommendation in its sample report is to let verified AI agents through the bot check on public pages and sign-up. That is often the right call — but bot checks exist because bots also commit fraud, scrape content and stuff credentials. Opening the gate to agents means deciding which agents, on which pages, with what rate limits, and keeping your fraud controls intact for everything else. A coding agent handed the fix kit will implement the instruction; a human still has to own the policy decision.
How Dedica Compares to Alternatives
Dedica is not the only tool chasing agent readiness, and the differences matter. AISaysYes, launched earlier in October, scores online stores specifically for AI shoppers, with free generators for agents.md and llms.txt files — a narrower, commerce-focused lens. AgentReady takes an API-toolkit approach, including a token-compression tool and an LLMO auditor, aimed at developers wiring agent readability into their stack. The open-source AI Visibility Checker runs 640+ checks per page, covering technical SEO, AI crawler access and agentic-web readiness, and can be run locally with no account at all.
Where Dedica wins is focus and finish. It does one thing — test whether agents can use your site — and it ends with executable fixes rather than a longer report. Where it loses is depth: the open-source checker covers far more surface, and a team that wants continuous monitoring or full-page crawling will outgrow a single scan quickly. Our read: start with Dedica’s free score to find your worst three failures, then decide whether you need a deeper tool. For most small businesses, the free score plus the $30 kit is the whole job.
From Our Experience: Fixing Agent Readiness in Practice
Having shipped websites for over nine years and more than 300 projects, we have watched this failure mode arrive gradually and then all at once. Three practical lessons stand out. First, server-render your money pages. If your pricing, plans or product details only appear after JavaScript runs, you are invisible to a large share of agents — and, incidentally, to some search crawlers and social previews too. Server-side rendering or static generation for pricing and product pages is the single highest-leverage fix most sites can make, and it helps humans on slow connections as well.
Second, treat agent access as a policy, not a toggle. Allowlisting verified agents on public and sign-up pages while keeping rate limits and fraud rules in place is the right shape, but it needs to be written down: which user agents, which paths, what limits, and who reviews the logs. We have seen teams flip a WAF rule off entirely to “fix” an agent issue and spend the next week fighting credential-stuffing traffic. Dedica flags the block; your team owns the rule.
Third, publish the machine-readable layer properly. An llms.txt file, an agents.md file, a public OpenAPI specification and, where it fits, an MCP server are how agents discover what you offer and what you allow. These files are cheap to create and they compound: every agent that reads them is one fewer agent guessing from your marketing copy. Dedica checks for their absence; building them is a half-day job for most sites and belongs in the same sprint as the other fixes.
When not to use Dedica’s fixes as-is: if your site handles payments, healthcare data or regulated content, run every access change past your security review first. The kit’s instructions are generic by design; your threat model is not.
Frequently asked questions
What exactly does a Dedica AI agent website check cover?
It runs 125 checks across four areas: discovery (can agents find and understand your site), access (can they reach your pages past bot checks), usability (can they complete journeys such as sign-up), and payments (can they finish a purchase). You get a score out of 100 and fixes ranked by impact.
Is the Dedica score really free?
Yes. The scan and score are free, take under a minute and require no sign-up. Only the fix kit is paid: $30 per site at launch pricing, $50 regular, as a one-time payment.
What is the most common reason websites fail the check?
In Dedica’s own examples, the top failures are bot checks blocking agents at sign-up, pricing pages that load their content only with JavaScript, and missing machine-readable documentation such as an OpenAPI specification or an MCP server. All three are fixable in a single sprint.
Does a high Dedica score guarantee AI agents will recommend my business?
No. The score measures whether agents can use your site, not whether they will choose you. Dedica’s claim that ready businesses are recommended 1.9 times more often is its own marketing figure, not independent research. Readiness removes a barrier; your product, pricing and reputation still have to win the recommendation.
Should I let AI agents past my bot checks?
Usually yes for public pages and sign-up, but as a written policy, not a blanket switch-off. Allow verified agents through on the paths they need, keep rate limits in place, and keep fraud controls active everywhere else. Review the change with your security team before shipping it.
The shift to agent visitors is not a future trend to watch; it is a traffic source already arriving at your sign-up page and leaving without a trace when it fails. A Dedica AI agent website check takes under a minute and tells you exactly where that silent failure happens — and the $30 fix kit turns the answer into work your developer or coding agent can execute. Run the free score, fix your top three issues, and rescan. If you want help turning the findings into a proper agent-readiness sprint — server-rendered money pages, a sane agent-access policy and the machine-readable layer done right — browse our AI Tools Directory for more tools we have tested, or talk to MTD Technologies about your site. Your next customer may not be a person at all.