The web was built for humans.
Nobody served the machines.

Every AI agent downloads 1.6 MB of noise — menus, scripts, ads, trackers — to extract 50 KB of content. Publishers lose bandwidth with zero revenue. No one knows who reads what.

93%Noise in HTML
$2.88To crawl 80 pages
1,400DOM elements/page
0€Revenue for publishers
Deploy CBOR-Web →
CBOR binary stream — structured, efficient, machine-native

55799({2:{name:"CBOR-Web"}})

⟳ Connecting to CBOR-Web node...

Machine-native copy of your site. No HTML parsing. No DOM traversal. 100× cheaper. 95% signal ratio.

100×
Cheaper
95%
Signal ratio
—
GitHub stars
38
Sites live

Docs — Get started

Add CBOR-Web to any site in 30 minutes. No CMS change.

📦 npm install

npm install @deltopide_edy/cborweb
Also: cargo, pip, composer, gem, go get. 8 languages, zero deps.

🔧 Convert

text2cbor generate -i ./html -o ./cbor -d mysite.com
Extracts headings, paragraphs, tables, images, Schema.org data.

🌐 Serve

Drop .cbor files alongside HTML. Rust server or Cloudflare Worker. Discovery via /.well-known/cbor-web.

🤖 SDK API

client.manifest() → full site. client.page(path) → single page. client.bundle() → one request. 8 languages.

🔐 Token gating

Agents hold ≥1 CBORW for premium access. Wallet-verified identity. Cryptographic signatures. Audit trail.

🏛️ IETF Draft

RFC 8949 · RFC 8610 · RFC 8615. Internet-Draft submitted. Open standard, CC BY 4.0.

🚀 Deploy in 2 commands

Copy-paste. 30 seconds. Your site becomes AI-readable.

1. Convert your HTML

text2cbor generate --input ./my-html-site --output ./cbor-output --domain mysite.com --default-lang enCopy

2. Serve the CBOR files

cbor-server --data ./cbor-output --listen 0.0.0.0:3001Copy

3. Verify

curl http://localhost:3001/.well-known/cbor-webCopy

That's it. AI agents now discover your content automatically. Humans see zero difference.

Blog

🆕 v2.2 Released

April 2026 — 8 SDKs, WordPress, Shopify, OCR pipeline, anti-DoS, npm, IETF draft.

View on GitHub

📊 93% token waste

200 websites benchmarked. Average: 120KB HTML → 8KB useful content. CBOR-Web: 95% signal ratio.

🌍 8 SDKs

TypeScript, Python, Go, PHP, Ruby, Java, C++, React. Zero deps. One API. npm published.

🏛️ Digital identity law

France requires AI agents to identify. CBOR-Web has wallet identity built-in since day one.

FAQ

🏆 Achievements

Earn badges contributing to CBOR-Web.

📂 GitHub — Live repository

Documentation pulled directly from the CBOR-Web repository.

📄 README.md

Loading from GitHub...

📋 Specs (7 docs)

CBOR-WEB-CORE · MULTIMEDIA · GENERATIVE · SECURITY · ECONOMICS · REFERENCE · DOLEANCE

Browse on GitHub

🛠️ Tools

text2cbor (Rust) · cbor-crawl (Rust) · cbor-server (Rust) · cbor-vectors · worker.js (Cloudflare) · benchmark.sh · OCR pipeline

View tools

📦 SDKs (8 languages)

TypeScript · Python · Go · PHP · Ruby · Java · C++ · React · WordPress plugin · Shopify snippet

View SDKs

🏛️ IETF Draft

Internet-Draft submitted. RFC 8949 (CBOR) · RFC 8610 (CDDL) · RFC 8615 (Well-Known URIs)

Read Draft

🔑 API & Token Access

Get your API key, token, and programmatic access to CBOR-Web services.

1. Create your account

Register on the text2cbor SaaS. You'll get an API key immediately. The first 5 conversions are free — then invoice-based billing (49€/site).

# Coming soon at api.cborweb.com curl -X POST https://api.cborweb.com/register \ -H "Content-Type: application/json" \ -d '{"email":"you@example.com","password":"..."}'Copy

2. Get your API key

Once registered, your API key is available at /api/key. Use it in the Authorization header for all API calls.

curl https://api.cborweb.com/api/key \ -H "Cookie: session=YOUR_SESSION_TOKEN"Copy

3. Upload HTML → CBOR

Send your HTML site as a zip file. Get back a zip of .cbor files ready to deploy.

curl -X POST https://api.cborweb.com/api/upload \ -H "Authorization: Bearer YOUR_API_KEY" \ -F "domain=mysite.com" \ -F "file=@mysite.zip"Copy

4. CBORW Token (Hold-to-Access)

Agents hold ≥1 CBORW token to access premium L1 content. Tokens are verified on-chain via verifyAccess(address). When an agent visits your site with X-CBOR-Web-Wallet header, the server checks their balance.

curl https://api.cborweb.com/verify-access?address=0x...Copy

Token contract: 100M CBORW · ERC-20 · Hold-to-access · 12-month founder cliff

📊 Token stats (live)

Token server: offline (prototype mode). Available on MiniPC/MacPro at port 3002.

🚀 Get Started — 3 steps

Convert your first site to CBOR-Web. Free. No credit card.

📧

1. Create account

📤

2. Upload HTML

Zip your HTML site and upload.

✅

3. Deploy

Download your .cbor files. Drop them alongside your HTML. Done.

GET /.well-known/cbor-web → 200 OK
Your site is now AI-readable ✨

5 conversions free · 49€/site after · Invoice-based

🔑 Your License Keys

Register to get your keys

API key for SaaS · License key for Desktop App

🔍 Web Audit v2.0 — Chrome · Firefox · Safari

Full site crawl, competitor analysis, backlinks, export PDF — powered by Rust/WASM.

📊 SEO Audit

Meta tags, headings structure, schema.org, sitemap, robots.txt, canonical URLs, alt texts, internal links.

🤖 GEO / AI Readiness

Checks if your site is optimized for AI agents. CBOR-Web detection, llms.txt, structured data quality, content-to-noise ratio.

📦 CBOR-Web Detection

Scans for /.well-known/cbor-web, verifies manifest format, checks content blocks, validates CBOR encoding.

🔄 Full Site Crawl

Follows all internal links, audits every page. Detects 404s, broken links, thin content, missing meta. Scores 0-100.

⚔️ Competitor Compare

Side-by-side with any competitor. Content gaps, keyword overlap, CBOR-Web advantage, score comparison.

🔗 Backlink Analysis

Estimated backlinks, indexed pages, domain authority signals. Full API integration for Ahrefs/Moz/SEMrush.

📄 Export Reports

Download audit as PDF, CSV, or HTML. Professional report with score, prioritized fixes, and recommendations.

🧩 Install the extension

1. Open Chrome → chrome://extensions 2. Enable "Developer mode" (top right) 3. Click "Load unpacked" 4. Select the clients/chrome-extension/ folder

Rust/WASM engine · Manifest V3 · Side Panel UI · Open source (MIT)

View on GitHub

📊 Live Metrics — Anonymous telemetry

Aggregated usage stats from CBOR-Web Audit Desktop users. Privacy-first — no personal data collected.

—
Sites scanned
—
Full crawls
—
Unique domains
—
Hours of audit

🔒 Privacy-first telemetry

We track only anonymized product metrics: scan count, crawl count, unique domains, usage time. No personal data, no URLs, no content. Telemetry is stored locally on your machine. Aggregate counts are shared only if you opt in via Settings.

⚙️ Manage your telemetry

Your local stats: —