Tools that check AI agent readiness: scanners and services
What each one looks at, in its own words, so you can pick by the question you need answered.
Reviewed September 2026
Several public scanners test whether a site's public pages work for agents, and several services handle part of the job, from bot control to fraud. We list them with what each checks, from its own pages. ora.ai's partners, Vercel and is-agentic.com, are listed in their own section.
Agent experience or AEO: which kind of tool is this?
AEO (answer engine optimization) platforms track and improve how your brand shows up in AI answers. That gets you surfaced: the find step. Agent experience is how agents experience your business once they arrive: can an agent read your site, get in, and complete a task such as a booking, a sign-up or a purchase. That gets you converted. ora.ai is an agent experience platform: its scan covers discovery, access, usability and payments, and its checks fit SaaS and brand sites. Scrunch, the closest comparison, focuses on how you appear in AI answers and also serves AI-optimized pages to agents. DataDome is a bot and agent trust platform: it decides which agents get in and blocks fraud, while ora.ai checks whether an agent that gets in can read your site and finish the task. TollBit is built mainly for publishers to control and charge for AI access; Axioma also works on the publisher side. Profound is listed under AEO platforms. See AEO vs agent experience for how agent experience is different from AEO platforms.
Scanners: enter a URL, get a report
Is It Agent Ready (isitagentready.com)
- Run by
- Cloudflare
- What it checks
- Five categories: discoverability (robots.txt, sitemap, Link headers, DNS for AI discovery), content accessibility (markdown content negotiation), bot access control (AI bot rules, Content Signals, Web Bot Auth), protocol discovery (MCP Server Card, Agent Skills, WebMCP, API Catalog, OAuth discovery, OAuth Protected Resource, Auth.md, ARD, Personal Agent Protocol (Poppy)) and commerce (x402, MPP, UCP, ACP).
- What you get
- A report by category. The page does not describe a score or levels.
- Limits it states
- Not stated on the page.
- Price
- Not stated on the page.
Source: isitagentready.com · Read 2026-10-18
ora.ai
- Run by
- ora.ai
- What it checks
- Four layers: Discovery (can an agent find and recommend you), Access (can it access your data and understand you), Usability (can it operate, authenticate, integrate) and Payments (can it pay you). Scan runs real agents against a site; Journey shows one agent's path on a page; a separate WebMCP audit reviews WebMCP tools in a real browser.
- What you get
- A grade from A+ (95–100, Leading) to F (0–27, Unusable), a report page per domain, and the same scan from a command-line tool (ax) and an MCP server.
- Limits it states
- Public URLs only; cannot scan private or internal sites or login-gated content.
- Price
- Not stated. The home page offers a waitlist; ora.ai's llms.txt says reads are open and need no key.
Sources: ora.ai · ora.ai methodology · ora.ai llms.txt · WebMCP audit · Read September 2026
Partners and built with Ora
Vercel is a partner of ora.ai, not a competitor. is-agentic.com was built by Vercel with ora.ai, every scan is run by ora.ai, and a result there is an ora.ai result. Vercel's bot management sits alongside a scan: it decides which verified agents get in. We list partners here, outside the scanners and services, and do not compare them with ora.ai.
is-agentic.com
- Run by
- Built by Vercel with ora.ai; scans are run by ora.ai (ora.ai partner)
- What it checks
- Essential checks such as server-rendered content, HTTP responses, document structure, error handling and usable controls; recommended checks that apply when a site offers APIs, OAuth, GraphQL, MCP servers, developer portals or commerce; bonus credit for emerging formats.
- What you get
- A score with evidence and recommendations. ora.ai's announcement gives essential checks 80 points, recommended 20 and bonus up to 5. Each report also includes an observed agent journey showing how one agent navigated the site and where it met friction, as supporting evidence. Completed reports are available through a public JSON API, an MCP server and a command-line tool (npx is-agentic).
- Limits it states
- Not stated on the page.
- Price
- Not stated on the page.
Sources: is-agentic.com · ora.ai announcement · Read September 2026
Vercel bot management
- Run by
- Vercel, an ora.ai partner
- What it checks
- An AI bots managed ruleset that can log or deny known AI crawlers, a directory of verified bots, and BotID fields that identify a verified bot's name and category.
- What you get
- Controls for sites hosted on Vercel.
- Limits it states
- Vercel says its bot protection does not work well behind a reverse proxy such as another CDN.
- Price
- Not stated on the pages read.
Sources: Vercel bot management · Handling verified bots · Read September 2026
Agent experience and readiness
Cloudflare AI Crawl Control
- Run by
- Cloudflare
- What it checks
- Which AI crawlers visit, what they request, and whether they follow your robots.txt; lets you allow or block them, and offers Pay per crawl in beta.
- What you get
- Controls and analytics in the Cloudflare dashboard.
- Limits it states
- For sites on Cloudflare.
- Price
- Available on all Cloudflare plans.
Source: Cloudflare AI Crawl Control · Read September 2026
Cloudflare Markdown for Agents
- Run by
- Cloudflare
- What it checks
- Not a checker: it converts HTML pages to markdown at the edge when an agent asks for markdown.
- What you get
- Markdown versions of pages, with a default content signal header if you set none.
- Limits it states
- Up to 6 MiB (6,291,456 bytes) of decompressed HTML.
- Price
- Pro, Business and Enterprise plans and SSL for SaaS, at no cost.
Source: Markdown for Agents · Read September 2026
Forter agentic commerce
- Run by
- Forter
- What it checks
- Forter Agentic Orchestration: headless agentic checkout across protocols through a single no-code partner integration, alongside its identity and fraud decisioning.
- What you get
- One integration connecting a store to agentic platforms, according to the page.
- Limits it states
- Not stated on the page.
- Price
- Not stated on the page.
Source: Forter · Read September 2026
nekuda
- Run by
- nekuda
- What it checks
- WebMCP tools on your site: AgentLane for managing them, a WebMCP Kit for coding agents, a Chrome extension to inspect, test and audit WebMCP tools, and an SDK.
- What you get
- Products for building and checking WebMCP tools.
- Limits it states
- Not stated on the page.
- Price
- Not stated on the page.
Source: nekuda · Read September 2026
webmcp.com
- Run by
- Maintained by nekuda
- What it checks
- Whether a site exposes working WebMCP tools; its scanner verifies the tools are live before listing a site.
- What you get
- A public directory, a read-only JSON API and a DevTools extension.
- Limits it states
- Sites with no registered tools are excluded.
- Price
- Not stated on the page.
Source: webmcp.com · Read September 2026
Closest comparison
Scrunch
- Run by
- Scrunch, a Sitecore company (acquired June 2026)
- What it checks
- How a brand appears in AI answers (prompt monitoring and citations), which AI bots visit the site (Agent Traffic), site audits that diagnose pages AI agents cannot retrieve or interpret (Site Optimization), and Agent Pages, which detects AI visits and serves AI-optimized versions of pages at the edge. For brands and agencies.
- What you get
- A free AI visibility audit (results in about 30 seconds, no credit card); dashboards, recommendations and page optimizations on paid plans.
- Limits it states
- The pages read do not describe running an agent through a task such as a booking or checkout, WebMCP, or agent payments. Agent Pages is listed on the Enterprise plan only.
- Price
- Core $250 a month with a 7-day free trial; Enterprise custom.
Sources: Scrunch · Agent Pages · Site Optimization · Pricing · Sitecore announcement · Read 2026-10-18
Bot and agent security: decide which agents get in
DataDome
- Run by
- DataDome
- What it checks
- Not a readiness checker. A bot and agent trust platform: it shows which bots and AI agents visit, verifies agents with Web Bot Auth, and lets a business allow, block, limit or charge them by name, down to a page or an action such as checkout, to stop fraud. Also protects MCP servers.
- What you get
- Agent traffic dashboards, access rules and fraud blocking; a free test of a domain's defenses against bots and untrusted agents.
- Limits it states
- Does not test whether an agent can read your site or complete a task.
- Price
- From $3,830 a month (Essentials); every plan asks for a demo.
Sources: DataDome · Agent Trust Management · Agentic commerce · MCP protection · Pricing · Read 2026-10-19
Publisher tools: control and charge for AI access
TollBit
- Run by
- TollBit (publisher-first)
- What it checks
- Not a readiness checker. Shows which AI bots and agents visit, lets a site choose what agents can access, and charges them through a bot paywall with usage-based pricing; also licensed access for AI retrieval and a separate Agent Site for agents. Built mainly for publishers; its site also lists brand and commerce offerings.
- What you get
- Analytics, access controls and paid access for AI visitors.
- Limits it states
- Does not score or test whether an agent can complete a task on your site.
- Price
- No public price list found; publishers are offered a free platform.
Sources: TollBit · Agent Site for brands · Read 2026-10-18
Axioma
- Run by
- Axioma (getaxioma.com), publisher-side
- What it checks
- Not a readiness checker. Insights into AI traffic, an AI-optimized version of the site for different agent types, and page-level business rules for AI access and monetization.
- What you get
- Access rules and AI traffic insights.
- Limits it states
- No scanner or agent task testing is described on the pages read.
- Price
- No public price; access by request.
Developer tools
514.ax
- Run by
- Fiveonefour
- What it checks
- For makers of developer tools: runs coding agents (Claude, Codex, Cursor and others) in clean sandboxes to see whether they find, choose and successfully use a product. Not a website scanner.
- What you get
- Agent transcripts, success rate, cost and time per variant, results on pull requests.
- Limits it states
- Aimed at developer products, not general business websites.
- Price
- Basic $0 a month plus usage ($25 usage credit included); Pro $250 a month plus usage; Enterprise custom; $0.30 per managed run.
AEO platforms
Profound
- Run by
- Profound
- What it checks
- The discovery side only: how a brand shows up in AI answers (Answer Engine Insights and Prompt Volumes across up to nine answer engines, including ChatGPT, Perplexity, Gemini, Microsoft Copilot and Claude), which AI bots crawl the site and AI-referred visits (Agent Analytics, from server logs), and a crawlability audit (bot classification, response time and rendering analysis). It describes itself as 'The AI marketing platform to win in' ChatGPT and AI search.
- What you get
- Visibility and AI crawler reports, page rewrites and new pages drafted for your approval, and a way to submit new content to AI crawlers.
- Limits it states
- The pages read do not mention checks for MCP, WebMCP or llms.txt, or running an agent through a task such as a booking or checkout.
- Price
- Free 7-day trial (50 prompts a day); Enterprise pricing is custom.
Sources: Profound · Profound Agent Analytics · Profound crawlability · Profound pricing · Read September 2026
Profound covers the discovery side: whether and how AI answers mention you, and which AI crawlers visit. It does not test whether an agent can then read, act or pay on your site, which is what the scanners above check.
Which one should I use?
Pick by the question. To see whether agents can find and read your public pages, either scanner above will do, and is-agentic.com (built with ora.ai) gives the same ora.ai scan for sites on Vercel. To see what happens when an agent tries a task, you need a tool that runs an agent: on this list, ora.ai says its Scan and Journey do. To see who is visiting, start with the analytics your CDN or host already provides. Automated tools work on public pages: ora.ai, for example, states that it cannot scan login-gated content.
Compare entries side by side
Put two to five of these entries next to each other, field by field.
Common questions
Is DataDome the same as ora.ai?
What is the difference between agent experience and AEO?
Is Scrunch the same as ora.ai?
Do any of these tools test with a real agent?
Is is-agentic.com a competitor of ora.ai?
Is Vercel a competitor of ora.ai?
Which of these tools publish a price?
Entries read on the dates shown with their sources; undated entries read September 2026. Partner section updated 2026-10-13. Profound entry read 2026-10-13. Tools page updated 2026-10-19.