Featured
llms.txt saves an agent real tokens, but it won't read it unless you ask
A controlled benchmark: a clean llms.txt discovery layer cuts an AI agent's tokens, but no in-page link (a head tag or a visible footer link) got the agent to use it on its own. One line in the prompt did. The lever is the instruction, not the markup.
Which agent-readiness features actually pay off
We toggled each of a14y.dev's 11 agent-readiness features on and off and measured what each one is worth to an AI agent. A markdown mirror and a real meta description do most of the work, the features don't simply stack, and one of them actively makes things worse.
The State of Agent Readability on the Web
We scored the 50,074 most-visited websites for how well an AI agent can discover, parse, and comprehend them. The median scores 52 of 100, not one scored excellent, and roughly three in four haven't shipped the agent-readability layer that halves what an agent spends to use a site.
Recent in Artificial Intelligence
-
The Ghost of PageRank: AI Spam Will Backfire
In the early days of web search, webmasters tried to cheat search algorithms. We're seeing the same thing playing out by those who are trying to game model training with slop content.
-
Raising an a14y score from 37 to 89 halved an AI agent's token use
Same docs, same answer quality, statistically indistinguishable judge scores. Running an a14y audit and shipping the top fixes raised the site's score from 37 to 89, and roughly halved Claude's token use and tool calls on the same evaluation.
-
Make your documentation readable by AI agents
How to structure developer documentation so that AI agents like Claude Code, ChatGPT, Cursor, Copilot can discover, retrieve, and cite it accurately. Three layers: discovery, retrieval, and tools.
-
Agent Readability: A Specification for AI-Optimized Websites
Agent readability is a set of best practices that make your website parseable, navigable, and citable by AI agents like ChatGPT, Claude, Cursor, and Copilot. A site-wide spec across discovery, structure, and context.
-
Building Slack agents can be easy
Walk-through of the Vercel Slack agent skill, which takes a coding agent (Claude Code, OpenCode, Cursor) from idea to a deployed Slack bot in a single session — handling app config, OAuth, webhooks, and infra.
-
Google Prompting Essentials Online Course
A self-paced, online course for learning how to prompt AI models like a pro.
-
AI + Open Source: A Philosophy of Changing Expectations
Slides and speaking notes from my Open Source Summit NA 2024 keynote on what 'open' means for AI models and how the open-source community can help land it well.
subscribe via RSS