journey
Follow the next step.
Inspect a recorded agent walk: what it found, where it stopped, and the evidence it left.
Explore Journey →agent readiness · waymark
Will an agent pick you?
Waymark measures public website readiness for AI agents using dated evidence across Discovery, Access, Usability, and Payments.
See what agents can find, open, and use on your site.
Featured scores
Selected examples from the public board.
Recent scores
Latest scans on the board · dates in UTC.
research
A view of the public board.
Explore the scores we have published. A changing sample of scanned sites, with dates and limitations.
Read the research →the api
A report your agent can read.
The same stored evidence, available as JSON, Markdown, and MCP.
Read the API docs →curl 'https://waymark.genedai.me/api/v1/report?url=vercel.com'Agent readiness FAQ
What does Waymark measure?
Waymark measures public agent readiness across Discovery (20), Access (30), Usability (40), and Payments (10). It checks signals such as llms.txt, readable content, and MCP discoverability.
Does Journey change the score?
No. An observed Journey is supporting evidence of one task and never changes the numeric score.
How do agents retrieve Markdown?
Send Accept: text/markdown to a report or human page, or use its linked .md twin. Completed reports also have a public JSON retrieve API.
Are MCP tools read-only?
The waymark_get_report, waymark_get_methodology, and waymark_get_developer_docs tools are read-only and require no key. The separate scan_domain tool can start or reuse a scan.
How are non-applicable payments treated?
Payments checks apply only when relevant payment or commerce evidence is present. N/A and scanner-error checks are excluded from the score denominator.
Does a score guarantee agent success?
No. A score describes a public snapshot at its scan time. Private pages, authenticated states, regional differences, and later site changes can affect results.