Framelink MCP for Figma
An MCP server that gives AI coding tools like Cursor read access to your Figma design files.GLips/Figma-Context-MCP
- Stars
- 16.1k
- Forks
- 1.3k
- Licence
- MIT
- Language
- TypeScript
- Last update
- 10 days ago
- Type
- CLI
What it is
This is a small server that sits between an AI coding assistant such as Cursor and the Figma API. You paste a Figma link into the chat and the assistant fetches the layout and styling details of that frame or group. The server trims the response down to the most relevant information before passing it to the model, which the README says helps the AI write more accurate code from a design.
- Paste a Figma frame link into Cursor agent mode and ask it to implement the design
- Have the assistant pull layout and styling metadata from a Figma group before writing code
- Run the server over stdio from an MCP client configuration file
Great for
- Developers using Cursor or another MCP-capable editor who want to turn Figma designs into code
- Front-end developers implementing a screen from a Figma frame in their chosen framework
- Teams that already have a Figma API access token and want the AI to read layout data instead of screenshots
Not ideal for
- People who do not use an MCP-compatible AI client, since the server is built to plug into one
- Anyone without a Figma access token, because the README says one is required
- Users who want a visual design tool rather than a code-generation aid
How to get it
- 1
Install with a package manager
You'll need: Node.js (version not specified in the README) · A Figma API access token · An MCP-compatible AI client such as Cursor
Run the server via npx (add to your MCP client config)
npx -y figma-developer-mcp --figma-api-key=YOUR-KEY --stdio - 2
Run it
- Create a Figma personal access token using the linked Figma help article
- Add the JSON snippet for your operating system to your MCP client configuration file
- Replace YOUR-KEY with your token, or set FIGMA_API_KEY in the env field instead
- Open Cursor's agent chat and paste a Figma file, frame, or group link
- Ask the assistant to implement the design
- 3
Get the most out of it
- Use the FIGMA_API_KEY and PORT environment variables in the env field instead of passing the key on the command line
- Link to a specific frame or group rather than a whole file to keep the context focused
- Consult the Framelink docs for further configuration details referenced in the README
Downloads
Strengths
- Shrinks Figma API responses to the layout and style details the model needs, per the README
- Simple setup: one JSON configuration entry using npx
- MIT licensed and open source
Trade-offs
- Requires creating and managing a Figma personal access token
- The README describes it as designed specifically for Cursor, so other clients are less clearly supported
- Output quality depends on the AI model and the design file's structure, which the README does not detail
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| Figma-Context-MCP (this) | 16.1k | CLI | beginner | 10 days ago | Developers using Cursor or another MCP-capable editor who want to turn Figma designs into code |
| Agent-Reach | 108k | CLI | beginner | 5 days ago | Developers building or using AI agents that need web content beyond basic search |
| codebase-memory-mcp | 48k | Dev tool | intermediate | 4 days ago | Developers who use an MCP-capable AI coding agent on a large codebase |
| agents | 42.1k | Dev tool | beginner | 4 days ago | Developers who already use one of the listed AI coding tools and want ready-made specialist agents |
| unity-mcp | 15.6k | App | intermediate | 7 days ago | Unity developers who want to prototype scenes by describing them in plain language |
| claude-context | 12.7k | Dev tool | beginner | 4 months ago | Developers working in large codebases where loading everything into an agent's context is impractical |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.