Agent Readiness Score
Scans a public URL and returns a 0-100 score based on what AI agents can discover, access, and use on the site.
I
Is Agentic
Alternatives
MAGL
Manus
Plans, executes, and delivers end-to-end work products.
AgentMail
Gives AI agents inboxes so they can send and act on email.
Grok Bot
Always-on AI teammates that get their own cloud computer, sign into your tools, and finish multi-step work on their own.
Liner
Meet AI agents purpose-built for professionals to enhance every workflow.
On this page
What is Is Agentic?
Is Agentic is a free tool built by Vercel on top of Ora's agent-readiness scoring engine. Enter any public URL and it scans what AI agents can discover, access, and understand on that site, then returns a 0-100 score plus a detailed report.
The score is applicability-aware: Essential checks (server-rendered content, correct HTTP behavior, clear document structure, recoverable errors, usable controls) carry most of the weight and apply to every site. Recommended checks only activate when a site actually offers an API, OAuth flow, GraphQL endpoint, MCP server, developer portal, or commerce surface — sites without those interfaces aren't penalized for lacking them. Emerging formats (like llms.txt) can add a small bonus but never lower a score if absent.
Each report also includes an observed agent journey — a real run showing how an agent navigated the site and where it hit friction — plus concrete, evidence-backed recommendations. Reports are publicly accessible at stable URLs, render server-side (so crawlers don't need JavaScript), and are available as Markdown or structured JSON, with a companion MCP server for programmatic retrieval. Is Agentic is powered by Ora and operated by Vercel.
Core Features
Essential + Recommended + Emerging Checks
Applicability-aware scoring: universal Essential checks, Recommended checks that only apply when relevant interfaces exist (API, OAuth, GraphQL, MCP, commerce), and small Emerging-format bonuses that never penalize absence.
Evidence-Backed Recommendations
Every check includes the evidence observed during the scan and, where applicable, a concrete fix that can be copied into an implementation workflow.
Observed Agent Journey
Each report shows a real agent's navigation path through the site and where it encountered friction, as supporting evidence alongside the numeric score.
Multi-format Report Access
Reports render as server-side HTML by default, are available as compact Markdown for agent consumption, and as structured JSON via a public API.
MCP Server
A read-only MCP server exposes completed reports as a tool with a declared schema, for use from MCP-compatible clients — no credentials or scan-triggering required.
Stable, Rescannable Report URLs
Each report lives at a persistent per-domain URL; rescanning replaces it with a fresh snapshot while keeping the same shareable link.
Screenshots
How to Use Is Agentic
- Go to is-agentic.com and enter the URL of the site you want to check.
- Start the scan — Ora crawls the public site and evaluates Essential, Recommended, and Emerging checks.
- Review the score (0-100) and the full report, which lists each check, the evidence found, and why it passed or failed.
- Fix the failed Essential checks first, since they cover baseline requirements any agent needs.
- Address Recommended gaps that match capabilities the site actually has (API, MCP server, commerce, etc.).
- Rescan the same URL after shipping fixes — the report refreshes in place at the same shareable link.
- Optionally pull completed reports programmatically via the public JSON API or the Is Agentic MCP server.
Use Cases
- Web teams and developers — audit a site before or after a redesign to confirm agents can still parse and use it correctly.
- Product and API teams — verify that documented APIs, OAuth flows, MCP servers, or GraphQL endpoints are actually discoverable and usable by agents.
- SEO and technical marketing — extend existing technical SEO practice to cover the growing share of traffic that comes from AI agents and assistants rather than human browsing.
- Ecommerce sites — check that commerce surfaces (product data, checkout flows) are structured in a way agentic shopping assistants can act on.
- Engineering leads — use the report as a prioritized technical checklist and re-run it after each fix to track progress over time.
- Developers building agents — pull structured JSON reports via the API or MCP server to programmatically decide whether a target site is safe/ready for an agent to operate on.
Pros & Cons
Pros
- Free to use, no login required to run a scan
- Applicability-aware scoring — sites aren't penalized for lacking interfaces (APIs, MCP, OAuth) they don't offer
- Every finding comes with concrete evidence from the actual scan, not a generic checklist
- Includes an observed agent journey showing real navigation friction, not just static checks
- Reports are stable, shareable URLs available as HTML, Markdown, or JSON
- Has its own MCP server so developers/agents can retrieve reports programmatically
- Backed by Vercel (interface, storage) and Ora (scoring engine), both established players in the space
Cons
- Scores only what's publicly visible — private pages, authenticated flows, and geo/bot-defense variations aren't captured
- One observed agent run is illustrative, not a statistically complete measure of every agent's behavior
- Not a certification or endorsement — doesn't cover security, accessibility, privacy, or legal compliance
- Usage is subject to published fair-use rules and rate limits, but the specific caps aren't disclosed on the pricing page
- Each report is a snapshot of one URL at one point in time — no built-in historical trend view, so tracking score changes across many domains means manually re-scanning and comparing reports
Pricing
Free
$0
- Public URL scanning and scoring
- Full report with Essential, Recommended, and Emerging checks
- Evidence-backed recommendations
- Observed agent journey included
- JSON API and MCP server access to completed reports
- No login or credit card required
- No paid plans, subscription charges, or per-report fees