System Architecture

How Onto Works

One edge layer. Two representations. Every AI agent gets clean, semantic content while humans see the full experience — zero code changes.

The Problem Today

AI agents are drowning in your React bloat

When GPTBot, ClaudeBot, or Perplexity crawls your Next.js site, they download hundreds of kilobytes of HTML, CSS framework classes, JavaScript bundles, and layout wrappers. Only 2-5% is actual semantic content.

596KB HTML downloaded → only 12KB of useful text
95%+ token budget wasted on Tailwind classes & SVGs
AI models hallucinate from noisy, unstructured input
WAFs block legitimate crawlers entirely
// What GPTBot receives today
<div class="flex min-h-screen flex-col
items-center justify-between p-24">
<div class="z-10 max-w-5xl w-full
items-center justify-between font-
mono text-sm lg:flex">
<!-- 400+ more lines of layout -->
<!-- Tailwind classes everywhere -->
<!-- Inline SVGs, tracking scripts -->
<p>Our product costs $49/mo</p>
<!-- The ONLY useful line -->
</div>
</div>
596KB total~0.8% signal
The Onto Pipeline

Three steps. Zero friction.

01

Request Arrives

Any visitor hits your URL. Onto's edge middleware inspects the Accept header and User-Agent in under 1ms.

Human browsers send Accept: text/html. AI agents send Accept: text/markdown or use known crawler User-Agents like GPTBot, ClaudeBot, PerplexityBot.

02

Edge Middleware Routes

Based on the signal, middleware rewrites the request to the right representation — HTML for humans, Markdown for agents.

No network calls for human traffic. Agent requests are rewritten to pre-compiled .onto payloads. Sub-15ms overhead in p99.

03

Clean Payload Served

AI agents receive compact, semantic Markdown/JSON with preserved hierarchy, tables, code blocks, and metadata.

10-100x smaller payloads. Zero hallucination noise. Accurate prices, descriptions, and documentation every time.

Architecture

Two paths. One URL.

Human Path
Browser RequestAccept: text/html
Middleware CheckNo agent signal → pass through
ResponseFull React/HTML experience
Payload596KB (unchanged)
Agent Path (With Onto)
Agent RequestAccept: text/markdown
Middleware CheckAgent detected → rewrite to .onto
ResponseClean Markdown + JSON-LD
Payload12KB (98% reduction)
Transformation

What Onto extracts

Onto Markdown Output
# Acme SaaS — Pricing
> Enterprise-grade automation for modern teams
## Plans
- **Starter**: $29/mo — 5 users, 10K events
- **Pro**: $99/mo — 25 users, 100K events
- **Enterprise**: Custom — unlimited everything
## Key Features
- Real-time analytics dashboard
- SOC 2 Type II certified
- 99.99% uptime SLA
12KB total100% signal
The Stack

Four components. One system.

LIVE

AIO Score Calculator

Free public tool. Enter any domain, get an instant AI-readiness report. The entry point to the Onto ecosystem.

ALPHA

@onto/next Extractor

Open-source Next.js plugin. Compiles your routes into clean Markdown at build time. npm install @onto/next.

ALPHA

Edge Middleware

Bot-aware traffic cop at the CDN edge. Routes AI agents to .onto payloads, humans to HTML. Sub-15ms overhead.

COMING

Control Plane (Pro)

SaaS dashboard for analytics, hidden prompts, and policy management. See which AI agents crawl your site and control what they see.

See it in action

Scan any website and see exactly what AI agents currently receive — and what they could get with Onto.