A few months ago we launched the Ora ranker, a score for how ready your product is for agents. It caught on faster than we planned. Every week more teams run their first scan, find they are further behind than they assumed, and start climbing.
That momentum is why we partnered with Vercel on is-agentic.com: an extension of the Ora ranker we built together, tuned for what most websites need. Enter a URL, get a score and the fixes to raise it.
Vercel is a leader in infrastructure for agents, and it shows on their own domain: vercel.com scores 85 on the Ora ranker, second of 310 companies in Infrastructure & DevOps, in the top 1% of the 16,000+ domains we scan. This is a partner that holds itself to the standard it is now putting in front of its builders.
One ranker, a verticalized endpoint
The Ora ranker measures the full surface a product can offer agents, 127 checks across four layers: discovery, access, usability, and payments. If agents can transact with you, through an API, an MCP server, or a checkout, the ranker measures all of it and ranks you against everyone else. A relevance pass sets aside what genuinely does not apply, so a free tool is never marked down for missing payment rails.
Agent readiness checkers are appearing everywhere, most of them a static checklist with a badge on top. A score is worth trusting when the methodology is public, the checks run live, and the number is ranked against everyone else. That is the engine Vercel chose to build on.
Most websites do not need half of that catalog. A site whose job is making its content easy for agents to find, read, and use has no API to gate and no checkout to wire. is-agentic, the endpoint we built together with Vercel, is tailored for exactly that kind of site: a small set of essentials carries the score, and capability checks like MCP servers or payments sit in a recommended tier and in bonus signals that only add credit.
Weighting differs too. The Ora ranker scores each check by importance, how much it moves agents in live runs, and recalibrates as agent behavior shifts. is-agentic weighs the checks in a tier evenly, which suits the essentials: they are the fundamentals, steady by nature, while the ranker keeps moving as our research progresses.
A report shaped for its audience
The two present results for different readers. The Ora ranker breaks a score down by layer and check weight, for the team that owns an API surface. is-agentic answers the question its audience is asking, is my site agentic and what do I fix first: a verdict in one line, a few plain buckets, and a short list of fixes.
Each report also shows a view from our journey engine: it gives an agent a real task and reports where the run struggled. That view comes from Ora Journeys, our tool for watching an agent use your site at journey.ora.ai. is-agentic shows one run per report; the full tool goes deeper.
Built on the Ora API
The essentials score lives in the Ora API. We added one option, include=essentials, that computes the essentials reading from the same scan the ranker already runs. is-agentic reads it straight from there:
curl -X POST "https://ora.ai/api/scan?include=essentials" \
-H "content-type: application/json" \
-d '{"url": "your-domain.com"}'The response adds an essentials object, with the score, the tier totals, a score for each surface, and prioritized fixes, next to the standard result. One scan, one engine, two readings. The option is public, so you can pull the same score is-agentic shows.
Which to run
If your site's job is content agents can find and use, is-agentic gives you the essentials fast and in plain language. If you own a product agents transact with, the Ora ranker is the universal standard. The two numbers will differ, and that's by design: one scan, two readings, each tuned to its audience. Same engine underneath, so a fix on one lifts the other.
Check your site on is-agentic.com. Put your product on the full ranker at ora.ai. Both are free, public, no signup.