AI-first by architecture, not bolt-on
Webflow and Framer added an "AI feature" on top of a SaaS canvas — there's no API to the design surface, so the AI is a chatbot bolted onto a black box. Meno is built the other way around. Your project is JSON files. AI reads files, edits files, ships files. The AI is in the loop end-to-end.
Why "file-based" matters for AI
LLMs are good at editing files. They are bad at clicking buttons in someone else's UI. Meno's data model is what AI is already optimized for.
Built-in AI assistant
Connect your Anthropic API key and chat with an AI that knows your project — your components, your design tokens, your CMS schema. Tell it what to change, see the diff on the canvas, accept or reject.
Works with Claude Code, Cursor, Copilot
Because the project is just files, any file-aware AI tool can edit it. Use Claude Code in the terminal. Use Cursor in your IDE. Use Copilot inline. The same AI you use for code edits your website.
Schema is shipped
Every Meno project ships with a CLAUDE.md file describing the schema, patterns, and conventions. Drop your project into any AI agent and it can produce valid output on the first try.
Visual canvas + AI prompt, together
Use the AI for the boring parts ("add a 3-column feature grid with these labels"). Use the canvas for the fiddly parts (drag this margin 4px). Use Git history when you change your mind. Three tools, one project.
Join teams that know the future is a hybrid of agents and humans
Work on your site in an agentic way without losing the visual editor.