Practical guides to agent-readiness
Understand what AI clients can discover, read and use on a website or API. Start with the buying decision, improve a technical surface or explore how AI visibility is measured.
Every guide here is re-read against its primary sources at least once a month, and the two families that move fastest, agent commerce and MCP discovery, more often than that. Specifications move, and a sentence that was true the day it shipped can stop being true without anything on this site changing. A layout change is not a technical re-check, so each guide's own check date is the one that counts.
Start here
A technical scan, manual review and observed AI answers reveal different problems.
How to choose an auditCompare an audit by its scope, evidence, deliverables and follow-up checks. See how the website and API audit differs from a focused Shopify check.
How technical readiness differs from AI visibilityThese labels describe overlapping work around discovery, answers and automated use.
Audit, visibility and priorities
What an audit measures, how to choose one, and how technical readiness differs from AI visibility.
A technical scan, manual review and observed AI answers reveal different problems.
How to choose an agent-readiness auditCompare an audit by its scope, evidence, deliverables and follow-up checks. See how the website and API audit differs from a focused Shopify check.
Make your website easier for AI assistants to find and citeClear, accessible content and consistent facts help retrieval systems understand a site.
SEO and agent-readiness: overlap and differencesSEO, AI-answer visibility and technical agent access overlap, but they answer different questions.
Agent-readiness, AEO and GEOThese labels describe overlapping work around discovery, answers and automated use.
Measure agent-readiness with evidenceA scan is one source of evidence. Combine it with direct technical checks and observed AI answers, and record the method and date behind every conclusion.
Common agent-readiness gaps in a measured sampleA prospecting sample of 567 company sites revealed recurring technical gaps.
Content and crawl access
What a text-based client can read on a site, and what the files at the root tell a crawler.
An llms.txt file offers a curated map of a site for clients that choose to read it. Learn its structure, discovery links and limits.
Serving Markdown to AI clientsA Markdown representation can make published content easier for text-based clients to consume. Keep its facts and links aligned with the HTML page.
Open Knowledge Format explainedOpen Knowledge Format describes knowledge as linked Markdown documents.
Sitemaps, robots.txt and AI crawler accessSitemaps describe published URLs, while robots.txt communicates crawler rules. Neither file guarantees retrieval or replaces access control.
Response headers for AI clientsResponse headers describe formats, discovery links and request policy. Check both what a header declares and whether the server behaves accordingly.
When AI clients cannot read rendered pagesSome clients read the initial HTTP response without running a browser.
JSON-LD and structured data for AI clientsStructured data makes page facts explicit, but those facts must agree with the visible page and underlying source.
Discovery and authentication
How a client finds an interface, and how the service decides what it may access.
A server card describes an MCP endpoint for clients that support the relevant discovery convention.
What agents.json describesagents.json is one pattern for describing actions and endpoints for automated clients.
The /.well-known directory for agent discoveryWell-known URLs give clients predictable places to look for specific metadata.
Agentic Resource Discovery and resource catalogsResource catalogs describe the agent-facing interfaces a site exposes.
Authentication and authorisation for AI agentsAuthentication identifies a client, authorisation determines what it may access or do.
Commerce and agent operations
Payment flows, commerce discovery, and the limits an agent operates within.
x402 describes an HTTP-based payment flow. A payment declaration, an accepted payment and settlement are different states and should be documented separately.
Agent commerce discovery: A2A, AP2, ACP and UCPCommerce discovery describes the interfaces and payment-related capabilities a service supports.
Agentic commerce readinessAgent commerce involves product information, permitted actions and a supported checkout path.
Define what an agent may do with your dataReliable agent operations depend on usable inputs, explicit permissions and observable outcomes.
AI agent use cases and their operating limitsExplore examples of agents reading data and taking permitted actions.
Frequently asked
What is an agent-readiness audit?
An agent-readiness audit measures how well an AI agent can discover, read, and act on a website or an API. It combines an independent scanner's per-check readings, a manual review of the agent-facing surfaces and observed answers from AI assistants to a documented question set, rather than a self-assessment.
Do I need llms.txt on my site?
llms.txt gives a curated map of what matters to the clients that fetch it, and no assistant is obliged to be one of them. It does not replace robots.txt or a sitemap, it complements them.
How do I get my site cited by AI assistants?
A model cites content it can read cleanly and corroborate. That means machine-readable surfaces such as llms.txt and structured data, a markdown form that does not exhaust the token budget, and being indexed where the assistant searches. None of that guarantees a citation.
What is an MCP server card?
An MCP server card is a JSON file that lets an agent discover a site's Model Context Protocol server, its endpoint and its transport, so the agent can connect without a human wiring up the connection. Deployed cards commonly sit at /.well-known/mcp/server-card.json.
Is agent-readiness the same as SEO?
No. SEO makes a site rank for a person to click. Agent-readiness makes a site legible and usable by an agent that reads and acts. A site can rank well and still be opaque to agents.
How is agent-readiness measured?
By an independent scanner that reads the live site and reports its checks one by one, beside a manual review and observed AI answers. A follow-up repeats the relevant checks and records the result beside the baseline, with the date and any change in the measurement method. No particular score or AI mention is guaranteed.
For an audit, contact info@turva.dev.