figwright

Name: figwright
Rating: 82 (1 reviews)
Author: awdr74100

MCP Server

awdr74100/figwright

Figwright is an open-source, bidirectional Figma agent for MCP clients that connects an MCP server to a Figma plugin via a local WebSocket relay. It enables AI agents to both read high-fidelity design context and write directly back to the canvas without requiring a paid Figma tier.

496 25MITUpdated 2026-08-12

Great

View on GitHub Open source file

Overview

Figwright provides a robust bridge between AI model context protocol (MCP) clients and Figma. By combining a Node-based MCP server with 112 specialized tools, a local WebSocket relay, and a Vue 3-powered Figma plugin, the system allows agents to ground code generation in real design tokens, component maps, and icons. It is designed to be provider-first, letting LLMs generate code aligned with the user's specific tech stack while maintaining synchronization and handling design diffs.

Capabilities

▸Bidirectional communication between MCP clients and Figma
▸Local WebSocket relay for zero-cost, paid-tier-free integration
▸Exposes 112 tools covering reads, writes, and advanced grounding
▸Component, token, and icon mapping joined with user codebases
▸Stack detection and codebase-aware code generation prompts
▸Design diffing and version skew management for plugin compatibility

Best for

Generating framework-specific frontend code directly from Figma designs, Reading design tokens, component maps, and icons for context-aware coding, Writing and modifying elements on the Figma canvas programmatically using AI agents, Analyzing design differences and establishing design baselines

Works with

Claude Code

Evoa Score breakdown= Σ (score × weight)

Task usefulness20%100 → +20.0

実タスクにどれだけ役立つか（機能の豊富さ・用途の明確さ）。 — AIによるcapabilities/use-cases解析

Code quality15%90 → +13.5

実装・指示の品質。 — AIによるSKILL.md/README解析

Maintenance15%100 → +15.0

リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ

Documentation12%49 → +5.9

ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量

Security15%80 → +12.0

危険・不審な挙動が無いか。 — AIによるセキュリティレビュー

Originality10%85 → +8.5

ありふれたラッパーではない独自性。 — AIによる独自性判定

Popularity8%68 → +5.4

コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）

Compatibility5%25 → +1.3

対応AIエージェントの広さ。 — AIによる対応エージェント判定

Weighted total81.6 / 100

ライセンス不明/制限あり（Red）のSkillは総合スコアに0.85倍の補正を適用します。ランキングはこのScoreのみで決まり、広告で変わりません。算出方法の詳細 →

Security considerations

The tool operates via a local WebSocket relay running on the user's machine, communicating between a local Node MCP process and a Figma plugin sandbox. No external cloud relays are utilized by default, keeping design data local.