Website checklist

Agent-ready website checklist for AI agent pages.

Use this checklist before training an AI agent on a client website. It turns crawlability, content depth, trust, and lead capture into a simple readiness score.

Published by Agent One Team.

Why this checklist exists

Checker demand

936 pageviews and 764 users in the last 90 days

Search demand

819 search-engine pageviews on the checker

Google entry point

816 checker pageviews from Google referrers

Agency fit

Readiness audits are the first step in the retainer playbook

Scoring model

Score each section from 0 to 2.

Give each checklist section 0 for absent, 1 for partial, and 2 for ready. The goal is not a perfect website. The goal is enough usable context to launch one focused AI agent page safely.

Score

0-5

Not agent-ready

Fix crawlability and missing service context first.

6-9

Trainable with gaps

Launch only after narrowing to one workflow.

10-12

Ready for a first agent page

Build, publish, and report visitor questions.

The checklist

Six checks before training an agent.

01

Crawlable page text

The content an agent needs should be present in HTML text, not only hidden in images, scripts, PDFs, or inaccessible app states.

  • Homepage, service pages, FAQs, policies, and contact details render usable text.
  • Navigation clutter does not overwhelm the useful page body.
  • Robots settings do not block the pages the agent needs to read.
02

Offer and service context

The website should explain what the business sells, who it helps, where it operates, and what a visitor should do next.

  • Core services, products, locations, and eligibility rules are explicit.
  • Pricing, quote, consultation, or booking expectations are explained where possible.
  • The first agent page can map to one clear buyer workflow.
03

FAQs and objections

Good agent pages need answers to the questions visitors already ask before they contact a business.

  • Service pages answer common what, how, when, cost, and policy questions.
  • Objections such as timing, safety, coverage, refunds, and requirements are documented.
  • Answers are specific enough for an AI agent to quote or summarize accurately.
04

Trust and proof

Agents need corroborating details so they can answer with confidence and route sensitive questions properly.

  • Reviews, case studies, credentials, years in business, or guarantees are easy to find.
  • Policies and limitations are visible enough to reduce hallucination risk.
  • Contact details, ownership, and business identity are consistent across the site.
05

Lead capture and handoff

Agent-ready content should make the next action obvious after the visitor gets an answer.

  • The right lead fields are known: name, contact, service, location, urgency, and context.
  • The routing destination is clear: email, Slack, CRM, calendar, inbox, or human review.
  • The agent has rules for when to capture a lead versus when to answer more questions.
06

Reporting loop

The site should be ready to improve from real visitor questions after the first agent page launches.

  • Unanswered questions can become content updates or new agent pages.
  • Qualified conversations can be reviewed without exposing sensitive raw data publicly.
  • Monthly reporting connects agent performance to the next client deliverable.

Agency deliverable

Turn the score into a client-ready audit.

The checklist is useful on its own, but it is strongest when paired with the live checker and a first-agent recommendation.

Readiness score and strongest gap

Checker output summary without raw private page text

Best first agent page recommendation

Lead capture fields and routing destination

Next content update for unanswered questions

Next step

Run the checklist against one page first.

A homepage audit is useful, but the first agent page should usually come from one high-intent service, quote, support, or appointment workflow.

FAQ

What makes a website agent-ready?

An agent-ready website exposes enough crawlable, structured, specific content for an AI agent to answer accurately and capture the right follow-up details.

Can an AI agent use a JavaScript-heavy website?

Sometimes, but important content should still be available as readable page text. If the checker returns missing or fragmented content, the site needs better semantic HTML or server-rendered fallback content.

How should agencies use this checklist?

Use it as the first client audit. Score the website, run the Agent-Ready Website Checker, identify the best first agent page, then package the findings into a readiness audit and implementation quote.

Should every page become an AI agent page?

No. Start with one high-intent workflow where visitor questions, lead capture, and content gaps are easy to measure.