Figma Motion MCP
Enables Claude to programmatically create and modify Figma animations (keyframes and motion styles) using the Figma Motion API via a plugin bridge.
README
Figma Motion MCP
Claude Code → MCP → our Figma plugin → Figma Plugin API. Authors Figma Motion
(keyframes / animation styles) programmatically from a prompt. No Figma agent, no use_figma.
Claude Code ──stdio(JSON-RPC)── MCP server ──ws://localhost:3055── plugin ui.html ──postMessage── plugin main thread ── figma.* (Motion API, Beta)
The MCP server process also hosts the WebSocket bridge, so there is no separate relay.
What it can do
| MCP tool | Plugin command | Purpose |
|---|---|---|
get_selection |
get_selection |
selected node id/name/type |
probe |
probe |
confirm Motion (Beta) API is exposed in your build |
read_motion |
read_motion |
current timelines / animated props / tracks / styles |
list_motion_templates |
(server local) | built-in code templates |
apply_motion_template |
apply_tracks |
resolve template+params → apply keyframe tracks |
apply_keyframes |
apply_tracks |
raw keyframe tracks |
set_timeline_duration |
set_timeline_duration |
timeline length (s) |
remove_motion |
remove_motion |
delete tracks and/or applied styles |
list_figma_animation_styles |
list_animation_styles |
native preset styles |
apply_animation_style |
apply_animation_style |
apply a native preset |
Create / modify / delete are all covered. Template resolution runs server-side — the plugin only ever receives concrete tracks, so it stays dumb and templates are versioned here.
Setup (teammates: one command)
The folder is self-contained (no monorepo deps). Hand someone this folder, they run:
cd figma-motion-mcp
./setup.sh
setup.sh does npm install, resolves your absolute node path (nvm-safe), and registers
the MCP server figma-motion at local scope (this project only, not committed — so
each teammate registers their own; no shared/hardcoded paths). It prints the remaining
manual steps.
Manual equivalent (if you prefer / no claude CLI)
npm install
claude mcp add figma-motion -- "$(command -v node)" "$(pwd)/server/mcp-server.js"
claude mcp get figma-motion # verify
nvm users: the absolute node path matters — the MCP subprocess won't inherit your shell PATH.
setup.sh/ the$(command -v node)form handles this automatically.
Then: run the plugin in Figma
Figma desktop → Plugins → Development → Import plugin from manifest… → pick
figma-motion-mcp/plugin/manifest.json. Open a NEW Claude session (it auto-starts the
server + WS bridge); the panel should turn green "Connected — bridge ready".
First run — runtime probe (do this once)
Typings (@figma/plugin-typings@1.130.0) confirm the Motion API contract, but it is Beta
and gated per build/seat. Before relying on it:
- In Figma, select any node, then run the plugin.
- In Claude: "run probe".
- Confirm
applyManualKeyframeTrack: "function"andfigma.motionis non-null. animatedPropertiesshows the real field names for your build — if any template name mismatches, fix it inserver/templates.js.
If applyManualKeyframeTrack is "undefined", Motion is not enabled in your Figma build yet.
Example flow
1) Select a frame/layer in Figma.
2) "get_selection" → nodeId
3) "apply fadeInUp to <nodeId>, duration 0.5" → apply_motion_template
4) "read_motion <nodeId>" → timelines / animatedProperties
5) "nudge TRANSLATION_Y keyframe to 32→0" → apply_keyframes
6) "remove OPACITY motion from <nodeId>" → remove_motion
Notes / limits
- Motion API is Beta — signatures may change. Write calls are wrapped; typings pinned to 1.130.0.
- No native style publishing (apply-only). Reuse lives in
server/templates.js. - Single-client MVP — newest plugin connection wins.
- No
SCALEfield — useSCALE_XY(corrected vs the original draft spec). - stdout is JSON-RPC only — all server logs go to stderr.
- Override the bridge port with
FIGMA_MOTION_WS_PORT(default 3055; updateui.htmlURL to match).
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.