What is agent readiness and why does it matter in 2026?
Agent readiness measures whether autonomous AI agents can act on your website, not just index or cite it. Since April 2026 Cloudflare's public Agent Readiness score checks discoverability, content, bot access control, API/auth/MCP discovery and, optionally, commerce. Early adopters can gain a first-wave advantage in agent-driven purchase flows.
Which standards does the audit check?
16 signals in five categories: discoverability (robots.txt, sitemap.xml, HTTP Link headers), content (Markdown content negotiation), bot access control (AI bot rules under RFC 9309, Content Signals, Web Bot Auth), API/auth/MCP/skills (API Catalog RFC 9727, OAuth/OIDC discovery, MCP Server Card SEP-1649, Agent Skills index, WebMCP) and commerce (x402, UCP profile, ACP discovery). Plus extended signals: llms.txt, .well-known/mcp.json, A2A and NLWeb.
What is the Model Context Protocol (MCP)?
MCP is an open protocol from Anthropic (2024) that gives agents standardized, JSON-RPC-based access to tools and data. Instead of scraping HTML, an agent calls typed functions such as searchProducts() or bookAppointment(). Adobe released an MCP server for Adobe Commerce in 2026; Shopify and major SaaS platforms offer MCP access too.
What is WebMCP and how does it differ from MCP?
WebMCP is a browser-native variant of MCP (Chrome 146+) built on the navigator.modelContext API. Pages expose tools directly, and agents call typed functions instead of analyzing screenshots. Early figures attributed to Google cite around 89 percent fewer tokens than visual automation. It is still experimental, so the audit assesses feasibility.
What is x402 and what role does it play in agent commerce?
x402 is Coinbase's open protocol for stablecoin payments over HTTP. The server answers with HTTP 402 and payment instructions; the client signs the payment in a request header. Stripe added x402 support for USDC on Base in February 2026. For agent micropayments (pay-per-read, API calls) it is the emerging standard.
What is the Agentic Commerce Protocol (ACP)?
ACP, developed by OpenAI and Stripe, handles secure instant checkouts in AI assistants. Its delegated payment model keeps the card with the merchant; the agent only passes on a signed purchase authorization. For brands it is the lever for being purchasable inside ChatGPT and similar surfaces.
What is Web Bot Auth?
Web Bot Auth is Cloudflare's 2025 mechanism for cryptographic bot identification, now in IETF drafts. A crawler publishes its public key and signs its HTTP requests; the server verifies the identity. Trusted agents get in, unverified ones are throttled.
How does agent readiness differ from SEO and GEO?
SEO makes content indexable, GEO makes it citable for LLMs, and agent readiness makes functions executable for agents. SEO and GEO are about visibility, agent readiness is about action. SEO brings the user, GEO earns the AI citation, agent readiness lets agents transact.
Which use cases does agent readiness enable right away?
Four clusters: agent commerce (search, configuration, checkout via ChatGPT, Gemini or Perplexity), service bookings (healthcare, travel, beauty, consulting), B2B pipeline automation (demo requests, pricing configurators, procurement discovery) and customer service deflection (knowledge base, status queries, self-service).
How long does an audit take and what do I get?
3 to 4 weeks. Phase 1: inventory and Cloudflare baseline. Phase 2: in-depth check of the 16 signals plus extended standards. Phase 3: roadmap, investment estimate, executive briefing. You get a 16-point score report, technical specifications, a briefing deck and a 90/180/365-day backlog.
Which industries benefit most?
Highest leverage: e-commerce (ACP, UCP, x402), travel (MCP booking, ACP), B2B SaaS (MCP for trials and demos), healthcare (booking with YMYL compliance) and financial services (OAuth discovery, delegated authentication). Medium: publishers (llms.txt, Markdown) and local service providers (availability APIs).
What does an agent readiness audit cost?
It depends on site complexity, implementation depth and industry. Three models: pilot (one brand, one market), standard (multi-brand or multi-market with MCP feasibility study) and a continuous retainer (quarterly re-score plus implementation support). Terms are set in the scope call.
