Docklet
Log in Get started

Your AI made something useful. Docklet makes it usable.

Publish AI-generated pages, guides, demos, and mini-sites as live links with comments, analytics, permissions, and version history — in-place right from your AI tool.

01
Created something with AI…
But how do I publish it?
Publish from any Claude, ChatGPT, or other AI tool to Docklet — no export, no hosting setup.
02
Need feedback…
But where should people comment?
Send one link and let reviewers comment directly on the live page.
03
Need to change it…
But don't want another copy?
Update the same Docklet. Comments and versions stay on one URL.
04
Shared it with people…
But did anyone actually look?
Track views, engagement, and activity so you know what happened after sharing.
claude.ai — chat

you Build an onboarding microsite for our new pricing and publish it to Docklet.

claude

claude Published to ryan.docklet.io/q3-pricing live

you Make the testimonials section bigger and republish.

claude Updated v3 · same link · 2 open comments preserved

A real prompt-to-publish flow.
01 — What is a Docklet

A live link for what AI already built.

The page, guide, or demo is done in chat.
Docklet is where it actually gets used.

Pricing pages, onboarding guides, pitch decks, prototypes, dashboards — anything your AI generates as HTML. Docklet gives it a URL people can open, comment on, and come back to while you keep iterating.

Once published, you can

Step 1

Share immediately

Send one durable link to people — not files, screenshots, or another export to chase down.

Step 2

Collect feedback in context

Get inline comments on the live artifact immediately — right where the discussion belongs.

Step 3

Control who sees it

Manage public, private, team, or client access — without rebuilding the artifact or moving platforms.

Step 4

See whether it worked

Track views, interactions, drop-offs, and returning readers — engagement analytics out of the box.

Step 5

Evolve without starting over

Update the same Docklet with versioning in place — history, rollback, and comments preserved.

02 — The problem

AI builds the page in minutes. Getting it to people still takes hours.

The HTML is ready, then the real work starts: find somewhere to host it, send the right link, collect feedback in email or Slack, lose context, regenerate, send another link. Docklet replaces that handoff with one publish step from your AI tool.

The current workflow

  1. Generate the artifact
  2. Export and host it somewhere
  3. Feedback scattered everywhere
  4. Lose context along the way
  5. Regenerate — share another link
  6. Repeat

With Docklet

  1. Ask your AI to publish it
  2. It's live
03 — How it works

Publish, share, update — same link throughout.

  1. 01

    Generate & publish

    Docklet plugs into your AI assistant through MCP. You ask for an artifact, your AI builds it, and Docklet deploys it. No exports. No Git. No hosting setup. No copy-pasting files.

    you → Create a Docklet explaining our Q3 strategy.
    claude →  ryan.docklet.io/q3-strategy
  2. 02

    Share & collect feedback in place

    Send one link. Reviewers comment directly on the experience — inline, threaded, contextual. No screenshots. No "see slide 14". Visibility controls keep what's private private.

  3. 03

    Iterate in place. See engagement insights.

    Ask your AI to revise — the same Docklet evolves. The link doesn't change. Comments stay anchored. Versions are tracked. Analytics tell you who opened it, what they interacted with, and where they dropped.

    1,284 views
    62% read-through
    3m 40s avg. read time
    23 comments
    3 versions
04 — Features

What you need after the page is generated.

AI-tool agnostic, MCP-native

Works with Claude, ChatGPT, Perplexity, and any tool that speaks MCP. Ask your AI to publish — you stay in the same chat.

Inline, threaded comments

Highlight anything, anywhere — replies, resolutions, archive.

Versioning that just happens

Every regenerate is a version. Roll back any time.

Engagement analytics

Views, drop-offs, interactions, returning readers — all in one place.

Visibility controls

Public, signed-in, or allowlisted. Org-level sharing. Custom domains.

Enterprise-ready foundations

Organizations, private spaces, role-based permissions, SSO, audit trails, custom domains, and self-hosting with data residency.
Built for security teams who need governance, visibility, and control.

05 — Why it exists

AI solved creation. Distribution didn't keep up.

You ask for a pricing page,
onboarding guide, or demo.

AI delivers HTML
in one conversation.

Then you're stuck:
where does it live,
who can open it,
how do people leave feedback?

Docklet handles that part.

06 — Pricing

Free to start. Honest at every tier.

Free

$0 forever

For tinkering, side projects, and your first few Docklets.

  • 1 Space
  • 5 Docklets
  • Public & selected-user visibility
  • 1 MCP-connected AI tool
  • Basic analytics
  • Current version only
Start free

Team

$19 per user / month

For organizations standardizing on AI-generated experiences.

  • Everything in Pro
  • Unlimited shared spaces
  • Organizations & private spaces
  • SSO (Google, Okta, custom OIDC)
  • Role-based permissions
  • Audit logs
  • Workspace analytics
  • Up to 10 versions & rollback
Talk to us

Enterprise

Custom tailored

For regulated industries and scale-first teams.

  • Self-hosting & data residency
  • Dedicated success team
  • Custom MCP integrations
  • Procurement & security review support
  • SLA
Contact sales
07 — FAQ

Things people ask us first.

Do I have to leave my AI tool to use Docklet?

No. That's the whole point. Docklet connects through MCP, so your AI assistant gets a "publish to Docklet" capability. You stay where you create.

What if my AI tool doesn't speak MCP yet?

You can also drop generated HTML into the Docklet console manually, or use our personal access token API. As MCP rolls out across tools, you get the one-prompt flow for free.

Can I update an artifact without breaking the link?

Yes. The Docklet stays at the same URL forever. Each regenerate creates a tracked version, comments stay anchored, and you can roll back at any point.

Who can see my Docklets?

You decide per-artifact: public, any signed-in user, or a specific list of emails. Team and Enterprise plans add organization-wide spaces and SSO.

What about data privacy?

We minimize data, encrypt in transit and at rest, never train on your content, and offer self-hosting on Enterprise. See our privacy policy.

Is this just hosting?

No. Hosting is the boring part. Docklet adds collaboration, versioning, analytics, visibility, and a publishing protocol that AI tools speak natively.

Stop sending screenshots. Send a Docklet.

Create an account and publish your first Docklet in minutes.