Agents · Gemini CLI
UI design from Gemini CLI, on a board your team can edit
Updated
Gemini CLI works in your terminal, next to your code. With Saymock it can also design: it lays out screens on a shared board as HTML + Tailwind, your team reviews them in the browser, and Gemini turns the approved screens into code in your repository.
What Gemini gets
- A board instead of files nobody opens. Screens are frames on a multiplayer board that the whole team can see, comment on and edit by hand.
- Your design system, from your code. Gemini reads your Tailwind config, fonts and icons in the repository and uses them on the board, so screens look like your product.
-
Self-checks.
get_screenshot,check_layoutandcompare_to_referencelet Gemini see and measure what it drew. -
Back to code.
export_codegives React, Vue or HTML with the same Tailwind classes;get_specsdescribes screens for native apps.
Set it up
-
Add the server, in a terminal:
gemini mcp add --scope user --transport http saymock https://app.saymock.com/mcp -
On first use, Gemini CLI opens the browser: sign in to Saymock (or create a free account)
and approve.
/mcp auth saymocksigns in again. - Ask Gemini for a screen. No API key involved.
Things to ask
Design the onboarding of this app in Saymock: three phone screens using our Tailwind colors.
Recreate our dashboard in Saymock from the code, then propose a denser layout.
Build the “Billing” screen from Saymock in this project, reusing our components.
Questions
Does Saymock call Gemini or another model?
No. Gemini CLI does the work with your own Google account or key; Saymock stores and renders the board.
How much does it cost?
Agents are free and unlimited: Saymock bills human editors only, and Free covers 2 editors per workspace. See the pricing.
The sign-in does not open.
Run /mcp auth saymock in Gemini CLI, and check the URL is exactly https://app.saymock.com/mcp.
Other agents: Claude Code, Codex, Cursor, all of them. Starting from an existing app: screens from your code.