Web platforms have been designed for humans — dashboards, buttons, form fields. But the next wave of web tooling isn’t human-first. It’s machine-readable.

Wapka supports MCP — the Model Context Protocol. This means AI models can interact with your site directly. Not through a chat interface bolted on top. Through the same APIs and documentation that developers use.

What MCP enables

MCP is a protocol that allows AI models to connect with external tools and data sources. On Wapka, this means:

  • Read your site data. Ask “What are my most popular pages?” or “Summarize my recent blog posts.”
  • Modify content. “Update the homepage hero text.” “Add three new portfolio items.”
  • Run scripts. “Write a Lua script that sends a welcome email to new users, test it, and deploy it.”
  • Manage files. “Take this image, resize it for web, and add it to the gallery with alt text.”
  • Configure the site. “Add a contact form. Enable social login. Set up email notifications.”

The AI becomes an operator of the platform — not just a chat assistant.

Why markdown docs matter

Wapka’s documentation is written in Markdown. That might seem like a small detail. It’s not.

Markdown is AI-native. When an AI model has access to a directory of Markdown files, it can:

  • Understand the full API surface without scraping HTML
  • Parse code examples correctly
  • Generate contextually accurate code

Platforms with PDF documentation, video-only tutorials, or scattered wiki pages can’t offer this. The docs need to be readable by both humans and models. Wapka’s are.

Use cases

For non-technical creators:

“Add a blog section to my site with categories and tags. Write three sample posts about photography. Create a contact form. Change the color scheme to match this brand palette I am pasting.”

No coding required. The AI handles the implementation.

For developers:

“Write a Lua API endpoint that accepts Stripe webhooks, validates the signature, and updates user subscription status in the database. Deploy it to staging first, then production after I approve.”

The AI becomes your pair programmer — with direct access to the platform.

For businesses:

“Generate a monthly content calendar. Write and schedule 8 blog posts. Create landing pages for the spring campaign. Update product pages with new pricing.”

Routine work moves to automation. Human time moves to strategy.

Not a gimmick

AI integration is often a checkbox feature — a thin wrapper over an API that feels bolted on. Wapka’s MCP support is different because the platform was already designed around structured, accessible data. The documentation is native. The API is first-class. The sandboxed scripting engine already runs untrusted code safely — whether that code was written by a human or generated by a model.

This isn’t a “ChatGPT plugin.” It’s a platform built for a future where software is operated by both people and autonomous agents.

Continue reading: Why Lua? The architectural decision → Wapka: the complete platform review →