Does my site need to be rebuilt in Next.js for the Agentic Web architecture?

No. The Discovery and Comprehension layers can be added to any existing site: WordPress, Webflow, Squarespace, or static HTML. The MCP server can be deployed as a standalone API alongside your existing site. A rebuild is only needed if your site is a client-side single-page app that renders no HTML for crawlers.

Agentic Web

The four-layer agentic web.

Q: How is Agentic Web optimisation different from traditional SEO?

Traditional SEO optimises for keyword-matching algorithms. The Agentic Web is optimised for language models that read, reason about, and cite content, and for AI agents that take actions on behalf of users. LLMs prioritise direct answers in the first 30% of content, structured schema, and machine-readable endpoints over keyword density or backlink count.

Q: How much does an Agentic Web audit cost?

An Agentic Web audit starts from £1,500 (~$1,900) and takes approximately one week. It includes a full audit of your site's current agent-visibility, gap analysis across all four layers, a prioritised implementation plan, and setup of llms.txt, agents.md, and robots.txt. A full four-layer build starts from £5,000 (~$6,400) and takes 1–2 weeks.

Most websites are invisible to AI agents: they can't be found, understood, or acted upon. We built a four-layer architecture that fixes this, and we built it for our own site first.

Build this for my site

Live on p0stman.com now
Four layers, all live
From £1,500

(01) The problem

Most websites are invisible to agents.

Invisible to crawlersSingle-page apps render <div id="root">. There is no content for a crawler to read. LLMs see nothing. Agents see nothing.

Unstructured contentEven crawlable sites bury answers in prose. LLMs can't cite what they can't parse. No schema, no citations.

No agent interfaceHumans click buttons. Agents need tools. Without an MCP server or WebMCP, agents have no programmatic way in.

(02) The proof

It works. Here's what happened when it did.

ChatGPT cited us, unprompted

A clinic manager in Dubai asked ChatGPT about voice agents for appointment booking. ChatGPT cited p0stman.com/locations/dubai/, specifically the answer capsule at the top of that page, which directly answered her query.

She clicked through, read the page, and submitted a project enquiry the same session. The enquiry converted to a discovery call. Zero ad spend. Zero cold outreach. No SEO campaign. The architecture did the work.

The answer capsule on that page said: “p0stman builds AI voice agents for Dubai clinics and medical centres: handling appointment booking, patient follow-ups, and multilingual reception in Arabic and English.” That sentence matched her query closely enough for ChatGPT to surface it as the answer.

LLM traffic converts differently

LLM-referred visitors convert 4.4x better than organic search (Adobe, 2025). Bounce rate is 45% lower: they arrive with high intent, having already had the question answered by an AI before clicking.

We track LLM source in our own Supabase analytics: UTM source, referrer domain, and traffic category. ChatGPT automatically appends ?utm_source=chatgpt.com to links it cites, so attribution is captured even when the referrer header is stripped.

Most sites don't track this at all. Most sites don't even render content that LLMs can read. That's the gap this architecture closes.

(03) The architecture

Four layers, each building on the last. All four are live on this site.

Can agents find this site?

Before an agent can do anything, it needs to find and index the site. This layer is about discoverability.

llms.txt (opens in a new tab)Standard index file for LLMs: lists every page with descriptions and categories
agents.md (opens in a new tab)Tells agents what they can do on this site, with full JSON-RPC examples
context.md (opens in a new tab)Complete company context in a single URL, shareable to any AI conversation
sitemap.xml (opens in a new tab)145 URLs indexed; submitted to Bing via IndexNow on every deploy
robots.txt (opens in a new tab)16 AI crawlers explicitly allowed: GPTBot, ClaudeBot, TavilyBot, PerplexityBot and more
Server-side renderingNext.js App Router: every page renders real HTML, no <div id="root"> for crawlers to bounce off.

(04) Live right now

What an agent can do with p0stman right now.

No human in the loop, no form-filling. Pure agent-to-server communication.

Book a discovery callAn agent representing a client can call book_discovery_call with name, email, and project description. The request lands directly in our CRM.

Submit a project enquirysubmit_inquiry with a detailed brief. Same result as filling in the contact form: stored in Supabase, Paul gets notified.

Research our servicesget_services, get_portfolio, search_content: any agent doing research on AI studios can pull structured data without scraping HTML.

(05) The offer

We build this for clients. We built it on ourselves first.

Every component, the MCP server, the AI endpoints, the schema markup, the llms.txt, the IndexNow pipeline, is a repeatable pattern we apply to any existing website or new build.

Agentic Web AuditA full audit of current agent-visibility, gap analysis across all four layers, a prioritised implementation plan, and llms.txt, agents.md and robots.txt setup.From £1,5001 week Full Four-Layer BuildAll four layers implemented: an MCP server with your business tools, AI-readable JSON endpoints, an IndexNow pipeline for ongoing indexing, WebMCP browser registration, and an A2A agent endpoint with Gemini inference.From £5,0001–2 weeks

USD equivalents at time of writing: ~$1,900 · ~$6,400. Billed in GBP.

The discovery layer

AI assistants, agent search engines and reasoning models are replacing the query box. If your content isn't structured for them: not in their training data, not indexable by their crawlers, not callable by their agents, you're invisible to an audience that is already making purchasing decisions without a single click on Google.

ChatGPT
Claude
Gemini
Grok
Perplexity
Meta AI

This is not the future. It is already happening. ChatGPT cites sources in every response. Perplexity crawls and indexes in real time. Claude reads your llms.txt before every session. Grok mines X for brand signals. Gemini surfaces structured schema directly in Google results. The question is not whether these systems will find your business. It is whether they will find it accurately, trust it, and act on it. That is exactly what the four-layer architecture solves.

(06) Questions

The ones people always ask.

What does an Agentic Web build include?

A full four-layer build includes: Layer 1 (Discovery): llms.txt, agents.md, context.md, sitemap setup, robots.txt configuration, IndexNow integration for real-time search engine submission; Layer 2 (Comprehension): JSON-LD schema on every page, structured AI-readable REST endpoints (/api/ai/context, /api/ai/services, /api/ai/portfolio), and answer capsule implementation across content pages; Layer 3 (Action): a full MCP server with JSON-RPC 2.0 protocol, custom tools for your business actions (bookings, enquiries, search), and WebMCP browser registration for Chrome. Layer 4 (Agent-to-Agent / A2A): AgentCard at /.well-known/agent.json, A2A JSON-RPC task endpoint at /api/agent, Zero responds autonomously to other agents using Gemini inference with your business context injected. Delivered in 1–2 weeks from £5,000 (~$6,400).

Does my site need to be rebuilt in Next.js?

No. The Discovery and Comprehension layers can be added to any existing site regardless of framework. llms.txt, agents.md, and JSON-LD schema work on WordPress, Webflow, Squarespace, and any static HTML site. The MCP server (Layer 3) requires a server-side runtime: this can be deployed as a standalone API alongside your existing site without touching the frontend. The only scenario where a rebuild makes sense is if your site is a single-page app (React, Vue, Angular) that renders client-side only, and those are genuinely invisible to AI crawlers.

How is this different from regular SEO?

Traditional SEO optimises for keyword-matching algorithms in search engines. The Agentic Web is optimised for language models that read, reason about, and cite content, and for AI agents that take actions on behalf of users. The signals are different: LLMs prioritise direct answers in the first 30% of content, structured schema, and machine-readable endpoints. An LLM doesn't care about keyword density or backlink count. It cares whether the page directly answers the question being asked.

How do I know if my site is already being cited by AI?

The clearest signal is UTM attribution: ChatGPT automatically appends ?utm_source=chatgpt.com to links it cites. If you have any analytics at all, you can check for this parameter in your traffic. Perplexity sends a referrer header from perplexity.ai. Claude and other models strip referrers but are working on attribution standards. A simpler test: search for your core topic in ChatGPT or Perplexity and see if your site appears in citations.

Can any AI agent call the p0stman MCP server right now?

Yes. The MCP server at https://p0stman.com/api/mcp is live and publicly accessible. Any MCP-compatible client can POST a JSON-RPC 2.0 request to discover tools (method: tools/list) and call them (method: tools/call). No API key required. The available tools are: book_discovery_call, submit_inquiry, get_services, get_portfolio, and search_content. The discovery manifest is at https://p0stman.com/mcp.json.

What is WebMCP and when does it matter?

WebMCP is a W3C-proposed browser standard (navigator.modelContext) that lets websites register tools directly with AI models running in the browser. When a user has an AI assistant in their browser (Chrome 146+ with the flag enabled, and eventually mainstream), that assistant can discover and call the registered tools without any external server. p0stman registers all 5 MCP tools via WebMCP on every page load. The flag is currently opt-in but expected to become default, similar to how Web Bluetooth or Geolocation went from experimental to standard.

(07) Deep dives

Comprehensive guides on every layer of the agentic web stack.