Rams MCP · The full engine, now in your coding agent

Rams Action

Agents write your UI now. The Rams Action is the judge in that loop — every change scored 0 to 100, merges gated on critical design defects, fixes handed back as patches your agents apply themselves.

Add to a workflow Get a free API key

The loop, in four steps

Your agents write the UI

Claude Code, Codex, Cursor — they already produce most of the interface code in a modern repo. Fast, plausible, and unreviewed.

Rams judges every change

The engine reads the changed UI files and scores them 0–100 against 291 design rules — accessibility, hierarchy, motion, craft. Any confirmed critical caps the score at 59.

Your agents repair

Findings come back as structured JSON plus git-applyable patches. Your own agent step applies them — Rams never touches your code.

The gate holds the bar

fail-on: critical blocks the merge until the defect is fixed. The loop keeps its speed; the product keeps its standard.

One YAML block

- uses: actions/checkout@v4
  with: { fetch-depth: 0 }
- uses: rams-design/rams-action@v1
  with:
    api-key: ${{ secrets.RAMS_API_KEY }}
    fail-on: critical            # or score<80, or never

Or run it on a schedule with mode: since-last-run — it reviews only what changed since the last sweep and leaves findings plus patches in rams-review/ for your own agent step to apply.

What comes back

scoreThe 0–100 design score — a number CI can compare, chart, and gate on

critical-countConfirmed criticals. fail-on: critical blocks the merge while it’s above zero

rams-review/issues.jsonEvery finding, structured: severity, category, file:line, the fix

rams-review/patches/One git-applyable diff per unambiguous fix — your agent runs git apply, not a guessing game

Block bad design before it merges

One YAML block. Free key, 30 reviews a month, no card.

Add to a workflow The full agent loop