Nobody asked us if we wanted it changed. So every week I check what's actually happening, and what to do about it, whether you use the web or build it.
The web now has visitors who aren't human. AI agents read your website, cite it, and increasingly buy from it, acting for the people who used to visit themselves.
No Hacks checks what AI is actually doing to the web, website by website, and tells you which parts are good, which are bad, and which are pointless. The articles, weekly podcast, and newsletter stay plain and back every claim with receipts.
No Hacks
A Publication About
the Agentic Web
66
Articles
233
Episodes
Weekly
Articles + Episodes
Evergreen and reference
These articles get revised rather than archived. Start with the definitions, then the two references that track every new crawler and agentic browser as it ships.
Definition
The canonical definition, and the Machine-First Architecture frame that runs through everything else here.
Definition
The W3C standard that hands agents callable tools instead of a page to read. Default-on across Shopify storefronts and Cloudflare-fronted websites since August 2026.
Updated September 2026
Five functional categories, every verified user-agent string, and every published IP range. Revised as new crawlers ship.
Updated September 2026
Every agentic browser, automation framework, and enterprise API, from Chrome auto-browse to Claude for Chrome.
The five-part field guide
Part 01
The shift from ranking pages for people to optimising them for the agents reading on their behalf.
Part 02
How AI systems choose the sources they quote, and what makes a page one of them.
Part 03
MCP for tools, A2A for agent-to-agent work, NLWeb and AGENTS.md for discovery.
Part 04
The accessibility tree, semantic HTML, and why a page that looks fine to you can be unreadable to an agent.
Part 05
ACP, UCP, payment rails, and how an agent completes a purchase without ever loading your checkout page.
New articles, the latest podcast episode, and a few links worth keeping. Practical strategies for making your website work for AI agents and the humans using it.
Shopify signs eligible stores up for every new AI shopping channel by default. Leave it on, and read the shopper count in your Shopify admin as clicks from AI apps, not agent purchases.
Amazon cut Meta's Muse off from its store on 20 September with a Conditions of Use notice. Its own robots.txt blanket-refuses 99 named agents, and Muse is not one of them, because Meta's crawler documentation lists no user agent string for it.
Strip a website down for machines and you should get something a machine can use. The text-only versions being served to AI are the page with the buttons removed, and GEO, the discipline for getting cited in AI answers, has organised itself around the half that cannot act.
Siri AI runs on Google's Gemini, and for a website the interesting part is not the model. Apple now reaches your pages five ways, three of them new this month, two with no user agent and nothing you can set. Every agent, every control, and what each control actually covers.
Voice shopping was forecast at $40 billion by 2022 and did not arrive, so the fair question is whether agentic commerce is the same story with a new noun. It is not. Voice moved the input and agents move the decision, and four things follow from that which have no precedent in any previous attempt. The catch is that the part doing the deciding is not what is running yet.
Bot Preference Sync turns your Cloudflare bot settings into robots.txt entries, and Cloudflare says it will reach every plan and arrive on by default for new customers. The product is good. Policy is set per category, and Cloudflare's tracked bot list decides which crawlers belong to each one, so the per-company judgment many websites make cannot be expressed in it. The Training switch also decides which crawlers get blocked for failing four disclosure conditions Cloudflare has published.
