Agent protocol readiness
Grade C
Weighted score across discoverability, content accessibility, bot access, protocol discovery, and commerce.
Score
Check whether your site speaks the protocols AI agents expect — from Markdown negotiation to MCP discovery and agentic commerce.
Updated
Loading tool interface...
Agent protocol readiness
Weighted score across discoverability, content accessibility, bot access, protocol discovery, and commerce.
Score
80/100 · weight 15%
25/100 · weight 20%
33/100 · weight 15%
17/100 · weight 25%
0/100 · weight 25%
robots.txt returned 200 with 412 bytes.
Sitemap directive points to https://example.com/sitemap.xml.
Homepage did not send any Link: response headers.
Server returned text/html for Accept: text/markdown.
Response did not include Vary: Accept.
Server returned 200 instead of 406 for an unsupported Accept.
Markdown was not preferred despite q=1.0 weighting.
Only GPTBot is explicitly referenced. Add ClaudeBot, Google-Extended, PerplexityBot.
No Content-Signal directives found in robots.txt.
/.well-known/http-message-signature-directory returned 404.
No server card at /.well-known/mcp/server-card.json or /.well-known/mcp.json.
/.well-known/agent-skills/index.json returned 404.
Homepage HTML has no <form toolname=…> or WebMCP meta tag.
/.well-known/api-catalog returned 200 with application/linkset+json.
/.well-known/oauth-authorization-server returned 404.
/.well-known/oauth-protected-resource returned 404.
No 402 response or PAYMENT-REQUIRED header detected.
No ACP manifest discovered at /.well-known/agentic-commerce.
No ucp:scopes found in OAuth authorization server metadata.
No MPP advertisement at /.well-known/machine-payments.
The Agent Protocol Readiness Checker runs twenty-plus probes against a URL to see whether it is ready for modern AI agents. It covers discoverability, Markdown content negotiation, bot access control, emerging protocol discovery endpoints, and agentic commerce standards.
Why it matters: AI agents increasingly prefer sites that advertise capabilities through well-known endpoints and serve machine-friendly formats. Exposing these signals lets agents discover tools and make payments without scraping.
Markdown Negotiation: Checks that `Accept: text/markdown` returns `Content-Type: text/markdown; charset=utf-8`, that `Vary: Accept` is set, that unsupported Accept types return 406, and that q-values are honored.
Bot Access Control: Verifies explicit rules for GPTBot, ClaudeBot, Google-Extended, and PerplexityBot in robots.txt, Cloudflare Content Signals directives, and a Web Bot Auth key directory.
Protocol Discovery: Probes `/.well-known/` endpoints for the MCP Server Card (SEP-1649 / SEP-1960), the Agent Skills manifest, the API Catalog (RFC 9727), OAuth Authorization Server metadata (RFC 8414), and OAuth Protected Resource metadata (RFC 9728). Scans the homepage HTML for WebMCP tool declarations.
Agentic Commerce: Detects x402 (402 Payment Required), ACP (Agentic Commerce Protocol), UCP (Universal Commerce Protocol), and MPP (Machine Payments Protocol) signals so you can verify whether agents can transact with your service.
After checking agent protocol readiness, validate robots.txt AI bot rules and run the AI Readiness Checker for content-structure coverage.
Validates Accept: text/markdown, Vary: Accept, 406 rejection, and q-value preference exactly like acceptmarkdown.com.
Checks MCP Server Cards, Agent Skills, WebMCP, API Catalog (RFC 9727), and OAuth discovery in one pass.
Detects x402, ACP, UCP, and MPP so you know whether agents can transact with your service today.
Answers about Agent Protocol Readiness Checker
The Agent Protocol Readiness Checker works by analyzing your URL or input and returning an immediate, actionable report. Use the findings for fast SEO implementation and QA.
Yes, the Agent Protocol Readiness Checker is free to use with no signup required. You can run checks instantly without paywalls.
Incomplete results usually happen when a website blocks automated requests or relies heavily on client-side rendering. Remote access restrictions can also limit what the tool can fetch.
Yes, these results are designed for practical client SEO audits and implementation handoff. They help teams prioritize fixes and validate outcomes quickly.