For a long time, creating a great codelab wasn’t the hard part. The hard part was everything that happened after the AI generated it.

We watched creators draft tutorials in Cursor, refine them in Claude, collect examples from different tools, and then spend the next 20–30 minutes copying content back into Stepwik.

Not because they wanted to.

Because that was the only workflow available.

Along the way, formatting broke, sections went missing, edits got out of sync, and the momentum of creating was lost. That was the moment we realized the problem wasn’t content creation anymore. The problem was the gap between AI and publishing.

Today, we’re excited to close that gap with the launch of the Stepwik MCP Server.

Introducing the Stepwik MCP Server

Connect Cursor, Claude, VS Code, Windsurf, and other AI assistants directly to your Stepwik workspace using the Model Context Protocol (MCP).

What changes now?

Instead of copy‑pasting between chat and Studio, you can ask your AI assistant to:

✓ Create a Firebase GenKit codelab

✓ Update section 3 with the new API

✓ Turn this tutorial into a course

✓ Search public Stepwik examples

And the assistant can work with your Stepwik content directly.

Built from real creator feedback

We didn’t build MCP because it was trendy. We built it because creators told us they wanted to:

1. Stay in flow while writing: Keep the creative momentum instead of switching tools.

2. Publish faster: Move from draft to published content with less friction.

3. Let AI help maintain docs: Update tutorials and courses without repetitive manual edits.

4. Work from their editor: Manage Stepwik content from the tools they already use every day.

Security was non‑negotiable

✓ New content is created as private drafts

✓ Access is limited by your scopes and workspace permissions

✓ API keys can be rotated or revoked anytime

✓ Publishing is always a conscious action in Studio

Try it now and follow the Guide

https://stepwik.com/labs/announcing-stepwik-mcp-server-8s3war