# HonestBot

> Agent readiness, measured in public.

HonestBot measures whether an AI agent can find, understand, and use a website. We publish the measurement. We do not sell scores or placement.

HonestBot's Agent Readiness Score (ARS) measures what a site reliably exposes for agents to find, understand, and use, under a fixed scanner profile. It is not a probability that a specific agent will complete a specific task, and it does not measure popularity.

Every site scanned under one scanner version gets the same checks, fixed scoring rules, scanner profile and evidence. No evaluator model writes the score. A higher score means more of the measured public surface is agent-ready; it is not a prediction of agent outcomes. ARS gives no credit for workarounds a capable agent might find.

Task-based scores depend on the agent, model, tools, task choice and prompt wording, vary between runs, and are blocked by logins and payments, so they do not compare sites like for like. All 42 scored checks look at what a company publishes for its own domain: its site, its subdomains, its DNS records, a status page it links to, and spec files it points to on other hosts. AI-answer citations (AEO, GEO), brand awareness, model priors, search rankings and directory listings are not scored, because they depend on the model, the question and the date as much as the site.

Compare scores only within one scanner version. Under the current scanner, 1.69.0, a site at the top level of every scored check scores 100. Versions before 1.68.0 topped out at 93 (1.13 to 1.67) or 94 (1.11 to 1.12) for a site with every check resolved. Use the score on the site's own result page and read its confidence level. Weights are product judgment, not yet calibrated against agent outcomes, and about 55% of points are in developer API, authentication and API documentation checks. We have not yet published test-retest numbers or a study of whether scores track agent outcomes.

Methodology: https://honestbot.com/methodology/

## Start here

- [What is agent readiness](https://honestbot.com/features/define-agent-readiness/): definition
- [Agent Comprehension](https://honestbot.com/features/agent-comprehension/): feature
- [Methodology](https://honestbot.com/methodology/)
- [Check criteria reference](https://api.honestbot.com/check-reference/): checks and points for every published scanner version. Example: [scanner v1.70.4](https://api.honestbot.com/check-reference/v1.70.4)
- [Homepage](https://honestbot.com/): scan a site
- [About](https://honestbot.com/about/): who we are and how we make money
- [Practice: make your site readable by agents](https://honestbot.com/blog/make-your-site-readable-by-agents/)
- [What does an HonestBot scan show?](https://honestbot.com/features/): the score, dimensions, checks and Agent Comprehension on a result
- [Why agent-readiness scores disagree](https://honestbot.com/blog/why-agent-readiness-scores-disagree/): how to read our score next to other scanners
- [When to use llms.txt, ARD, and MCP together](https://honestbot.com/blog/when-to-use-llms-txt-ard-and-mcp/): which agent surfaces belong on a public product
- [Blog](https://honestbot.com/blog/): all Define, Measure, Practice and Rank posts

## Optional

- [llms-full.txt](https://honestbot.com/llms-full.txt): longer agent notes on measurement, crawl surfaces, and API pointers