Create pages, and edit copy from inside Cursor / CoWork / Copilot etc - no human involvement needed.
Humans get a beautiful online editor too.
One-line activation - paste into Cursor / Claude / etc
Up and running in under two minutes
Your AI reads the skill.md file and registers a free account using its MLAuth identity.
The agent automatically drops an API key into your project. No manual steps.
Pages, text blocks, images - the agent writes and updates content via the API.
Then just ask your AI…
Built for developers who want to ship fast without wrestling with bloated CMS platforms
Use our npm package or hit the REST API directly. Your choice, your way.
Give your team a beautiful editor on pullnote.com. No training required.
Agents authenticate with MLAuth cryptographic identities - no API keys to manage.
Generate placeholder content with AI. Let humans review and approve later.
Organise content with paths that match your site structure. Simple and intuitive.
Automatic sitemaps, metadata management, and image generation included.
No complicated setup. No config files. Dovetails into your SSR. Supports Next.js, Nuxt and SvelteKit.
Install the package
npm install @pullnote/client
Fetch your content
Get HTML or markdown for any path
Render on your page
Drop it inline on a page or template
npm install @pullnote/client
import { PullnoteClient } from '@pullnote/client';
const pn = new PullnoteClient(PULLNOTE_KEY );
// Store
await pn.add('/about', {
title: 'About',
content: 'About the **Pullnote** Headless Content API'
});
// Retrieve
var content = await pn.getMd('/about'); // For the raw Markdown
var content = await pn.getHtml('/about'); // If you prefer to directly render as HTML
Traditional CMSes can't be controlled by AI. Pullnote can.
AI agents authenticate with cryptographic identities - no passwords, no OAuth. Sign with a private key.
Pass your LLM a link to our llms.txt file and it can create, edit, and retrieve content on your behalf.
Append .md to any URL for clean markdown. Example: /docs/llms.md
Pullnote is built on MLAuth — a public-key authentication standard designed specifically for AI agents. No shared secrets, no passwords, no API keys sitting in .env files.
Agent generates a keypair
On first run the agent creates an asymmetric keypair locally. The private key never leaves the agent's environment.
Identity registered with MLAuth
The public key is registered with mlauth.ai, giving the agent a verifiable, globally-unique identity — no human account needed.
Requests are cryptographically signed
Every API call to Pullnote is signed with the agent's private key. The server verifies the signature against the public key on record.
Agents act on behalf of humans
Humans grant agents permission to edit specific projects. The agent's signed identity is the audit trail — every change is attributed and traceable.
Zero-trust by design
No shared secrets
Private keys stay with the agent. Pullnote only stores the public key.
No password reuse
Each agent identity is unique and cannot be impersonated without the private key.
Signed audit trail
Every content change is attributable to a verified agent or human identity.
Revocable access
Humans can revoke an agent's permissions at any time from the console.
Join developers who chose simplicity over complexity
See API reference on GitHub for all options