Scanners that test with real agents
Which scanners send a real agent at your site, and which only read files and headers.
In short
4 of the 10 scanners put a real agent on a site, and only Ora does it on demand with no account or key. StartupHub runs six agent frameworks on a free account, Scope runs coding agents on request, and is-agentic.com shows runs recorded by Ora.
Who runs a real agent
| Scanner | Points | What it does | Source |
|---|---|---|---|
| 15 of 15 | Journey runs a real agent on a curated task through POST /api/journey/runs with no key; custom tasks need a partner key | ora.ai/llms.txt | |
| 12 of 15 | Live Agent Run, three runs a month on a free account | startuphub.ai/agent-readiness | |
| is-agentic.com | 6 of 15 | shows observed agent journeys from Ora | is-agentic.com/methodology |
| 0 of 15 | does not send an agent to complete a task | agent-ready.dev/methodology | |
| 0 of 15 | fetches files, headers and DNS records; runs no agent | isitagentready.com/llms-full.txt | |
| 0 of 15 | none | none | |
| 0 of 15 | deterministic audits, no agent run | developer.chrome.com/docs/lighthouse/agentic-browsing/scoring | |
| SEOJuice Agent Ready | 0 of 15 | reads files, headers and markup | seojuice.com/tools/agent-ready |
| 12 of 15 | runs coding agents on repeatable tasks; access is requested | github.com/microsoft/scope | |
| 0 of 15 | none | none |
A scanner that fetches files, headers and DNS records can tell whether a signal is present. It cannot tell whether an agent that arrives finds the page useful, which is what a run on a task shows.
Ora Journey
- Journey points a real agent at a domain with an intent such as "find pricing" and records every step, stall and outcome.9
- POST /api/journey/runs runs a curated task anonymously; free-text tasks and POST /api/journey/domains/{host} need a partner key.10
- Each domain is capped at 100 journey runs per rolling 24 hours and cannot run two at once.10
Journey is the part of Ora the rubric rewards most under real agents: 15 of 15, because a curated task runs on demand with no key. The Journey guide says what a run shows and what it cannot.
StartupHub's Live Agent Run
StartupHub starts six agent frameworks on its own hardware and asks each the same seven questions about a site, eight when a way to pay is found. The page reports 1,307 runs on 217 sites since 1 October 2026, measured 22 October 2026; these are vendor-reported figures 2.
| Framework | Maker | Answers | Runs | Fetches | Seconds |
|---|---|---|---|---|---|
| Claude Code | Anthropic | 4.7 | 223 | 5.3 | 45 |
| Codex CLI | OpenAI | 4.8 | 223 | 6.1 | 32 |
| OpenCode | Open source | 4.8 | 223 | 5.4 | 20 |
| Cursor | Cursor | 5.2 | 209 | 9.0 | 63 |
| Gemini CLI | 6.1 | 25 | 7.3 | 163 | |
| Plain model with a fetch tool | Control | 2.9 | 404 | 1.3 | 2 |
What Ora's lab found about agent behaviour
These are Ora's own research findings, labelled as vendor claims with their dates; this site has not replicated them.
- Ora's lab found (19 Oct 2026) that of 19 configurations of one docs site, only hiding the answer in JavaScript (0 of 5 for plain fetch) and returning 403 to agent user agents (0 of 5 for both clients) stopped retrieval, over 190 probes.11
- Ora's lab found (19 Oct 2026) that agents reached docs pages in 83 percent of traced runs, the homepage in 69 percent, .well-known paths in 23 percent, openapi.json in 21 percent and sitemap.xml in 4 percent.11
Scanners that read signals only
agent-ready.dev, Cloudflare, Mintlify, Lighthouse, SEOJuice and SiteSpeakAI score 0 on this criterion. Each states what it does instead: for example agent-ready.dev says it does not send an agent to complete a task, and Lighthouse runs deterministic audits. That is a design choice, not a flaw.
Sources
- ora.ai/llms.txt. https://ora.ai/llms.txt, read 2026-11-05.
- startuphub.ai/agent-readiness. https://www.startuphub.ai/agent-readiness, read 2026-11-05.
- is-agentic.com/methodology. https://is-agentic.com/methodology, read 2026-11-05.
- agent-ready.dev/methodology. https://agent-ready.dev/methodology, read 2026-11-05.
- isitagentready.com/llms-full.txt. https://isitagentready.com/llms-full.txt, read 2026-11-05.
- developer.chrome.com/docs/lighthouse/agentic-browsing/scoring. https://developer.chrome.com/docs/lighthouse/agentic-browsing/scoring, read 2026-11-05.
- seojuice.com/tools/agent-ready. https://seojuice.com/tools/agent-ready/, read 2026-11-05.
- github.com/microsoft/scope. https://github.com/microsoft/scope, read 2026-11-05.
- Ora home page. https://ora.ai/, read 2026-11-04.
- Ora complete public documentation (llms-full.txt). https://ora.ai/llms-full.txt, read 2026-11-04.
- Introducing AgentReady.org (Ora blog, 2026-10-19). https://ora.ai/blog/agentready-launch, read 2026-11-04.
Last reviewed 2026-11-05