MCP.so
Sign In

RamsVerifiedFeatured

@HSLA0001 Inc.

About Rams

Design review for UI code, inside your agent. One tool (review_files): up to 20 files in, a scored review out — 0–100, issues with severity and file:line, a concrete fix for each. 258 versioned rules across accessibility, color, typography, spacing, components, UX, motion, craft, and native SwiftUI. Free API key (30 reviews/month, no card): https://www.rams.ai/agents

Connection details

https://worker.rams.ai/mcp

Setup

claude mcp add rams --transport http https://worker.rams.ai/mcp --header "Authorization: Bearer rams_YOUR_API_KEY"

Tools

No tools detected

Fetch the live tool list directly from this server's endpoint using the button above.

Overview

What is Rams?

Rams reviews UI code for design quality — the things linters can't see: weak hierarchy, hardcoded colors bypassing a design system, missing interaction states, motion without reduced-motion guards, inaccessible controls.

Your agent writes the code; Rams reviews it before it's committed.

The tool

review_files — send up to 20 UI files from the working tree, get back:

  • A 0–100 design score (any confirmed critical issue caps it at 59)
  • Issues with severity, category, and file:line
  • A concrete fix for every issue — not a lecture, a change

Also usage — check remaining quota, free to call.

What it checks

258 versioned rules across 9 categories: accessibility (WCAG 2.2 AA), color, typography, spacing, components, UX, motion, craft (AI-slop patterns), and native (SwiftUI / Apple HIG). Findings are verified server-side before they count.

Works with React, Vue, Svelte, Angular, React Native, SwiftUI, and HTML/CSS.

Setup

  1. Get a free API key at rams.ai/app/settings — 30 reviews/month, no card
  2. Add the server (config above) with your key
  3. Ask your agent for a design review before committing UI changes

Recommended standing instruction for AGENTS.md / CLAUDE.md:

Before committing UI changes, run a Rams design review (review_files) and fix any critical issues.

Same engine, three surfaces

The identical engine reviews every pull request via the GitHub App and scores public repos at rams.ai/scores — see a real review before you install anything: rams.ai/score/antiwork/gumroad

Frequently asked questions

What is the Rams remote MCP server?

The Rams remote MCP server is a hosted Model Context Protocol endpoint at https://worker.rams.ai/mcp, so AI assistants can connect to it without installing or running anything locally.

How do I connect to the Rams MCP server?

Add the endpoint https://worker.rams.ai/mcp to any MCP-compatible client such as Claude Code, Cursor, or VS Code. The setup snippets on this page configure each client in one step.

Does the Rams MCP server require authentication?

Yes. Rams requires an API key — check the official documentation for how to obtain one, then include it in your client's configuration.

Which transport does the Rams MCP server use?

Rams exposes a Streamable HTTP endpoint, the transport used by remote MCP servers and supported by all major MCP clients.

Comments