OpenPencil

Open-source design editor. Opens .fig and .pen design files, includes built-in AI, and ships as a programmable toolkit with a headless Vue SDK for building custom editors.

Status: Active development. Usable today, with some rough edges as features evolve.

Try it online → · Download · Documentation · Roadmap · llms.txt

Installation

macOS (Homebrew):

brew install --cask openpencil

Or download from the releases page, or use the web app — no install needed.

What it does

• Opens .fig and .pen files — read and write native Figma files, open supported Pencil documents from the app or OS file browser, copy & paste nodes between apps
• AI builds designs — describe what you want in chat, 90+ tools create and modify nodes. Connect OpenRouter, Anthropic, OpenAI, Google AI, Z.ai, MiniMax, or compatible endpoints
• Fully programmable — headless CLI, XPath queries, Figma Plugin API via eval, MCP server for AI agents, and desktop agent integrations for Claude Code, Codex, and Gemini CLI
• Lint, convert, and extract tokens — inspect documents, lint naming/layout/accessibility, convert between supported formats, analyze colors/typography/spacing/clusters, and extract design tokens
• Components and variants — create reusable components, group variants into component sets, insert local assets as instances, and switch variants from the inspector
• Image vectorization — convert image layers into editable vector layers with Recraft or fal.ai
• Design-to-code export — export selections as JSX/Tailwind, generate token outputs, and map designs into component-oriented code workflows
• Vue SDK for custom editors — headless components and composables for embedding OpenPencil into other apps or building workflow-specific editing surfaces. Read the SDK docs →
• Real-time collaboration — P2P via WebRTC, no server, no account. Cursors, presence, follow mode
• Auto layout & CSS Grid — flex and grid layout via Yoga WASM, with gap, padding, alignment, track sizing
• ~7 MB desktop app — Tauri v2 for macOS, Windows, Linux. Also runs in the browser as a PWA

CLI

npm install -g @open-pencil/cli
# or: bun add -g @open-pencil/cli

Inspect design files

Browse node trees, search by name or type, dig into properties — all without opening the editor:

openpencil tree design.fig
openpencil find design.pen --type TEXT
openpencil node design.fig --id 1:23
openpencil info design.fig

[0] [page] "Getting started" (0:46566)
  [0] [section] "" (0:46567)
    [0] [frame] "Body" (0:46568)
      [0] [frame] "Introduction" (0:46569)
        [0] [frame] "Introduction Card" (0:46570)
          [0] [frame] "Guidance" (0:46571)

Query with XPath

Use XPath selectors to find nodes by type, attributes, and structure:

openpencil query design.fig "//FRAME"                              # All frames
openpencil query design.fig "//FRAME[@width < 300]"                # Frames under 300px
openpencil query design.fig "//TEXT[contains(@name, 'Button')]"     # Text with 'Button' in name
openpencil query design.fig "//*[@cornerRadius > 0]"               # Rounded corners
openpencil query design.fig "//SECTION//TEXT"                       # Text inside sections

Export

Render to PNG, JPG, WEBP, SVG, .fig, or JSX — or export selections/pages as .fig and convert whole documents between supported formats:

openpencil export design.fig                           # PNG
openpencil export design.fig -f jpg -s 2 -q 90        # JPG at 2x, quality 90
openpencil export design.fig -f fig --page "Page 1"   # Export a page as .fig
openpencil export design.fig -f jsx --style tailwind   # Tailwind JSX
openpencil export design.fig -f html --css tailwind    # Tailwind HTML fragment
openpencil export design.fig -f html --html standalone --assets external # HTML + assets
openpencil convert design.pen output.fig               # Convert between document formats
openpencil import page.html --css styles.css -o page.fig # HTML/CSS → editable .fig

DOM/CSS input flows through @open-pencil/dom-css, so HTML, authored CSS, and Tailwind utility CSS can become editable OpenPencil layers:

openpencil import card.html --css card.css -o card.fig
openpencil import card.html --tailwind "flex flex-col gap-3 w-80 p-6 rounded-xl bg-white" -o card.fig

<div className="flex flex-col gap-4 p-6 bg-white rounded-xl">
  <p className="text-2xl font-bold text-[#1D1B20]">Card Title</p>
  <p className="text-sm text-[#49454F]">Description text</p>
</div>

Lint design files

Catch naming, layout, structure, and accessibility issues from the terminal:

openpencil lint design.fig
openpencil lint design.pen --preset strict
openpencil lint design.fig --rule color-contrast
openpencil lint design.fig --list-rules

Analyze and extract design tokens

Audit an entire design system from the terminal — find inconsistencies, extract the real palette, and spot components waiting to be extracted:

openpencil analyze colors design.fig
openpencil analyze typography design.fig
openpencil analyze spacing design.fig
openpencil analyze clusters design.fig
openpencil analyze overlaps design.fig
openpencil variables design.fig

#1d1b20  ██████████████████████████████ 17155×
#49454f  ██████████████████████████████ 9814×
#ffffff  ██████████████████████████████ 8620×
#6750a4  ██████████████████████████████ 3967×

3771× frame "container" (100% match)
     size: 40×40, structure: Frame > [Frame]

2982× instance "Checkboxes" (100% match)
     size: 48×48, structure: Instance > [Frame]

Script with Figma Plugin API

eval gives you the full Figma Plugin API. Modify the file, write it back:

openpencil eval design.fig -c "figma.currentPage.children.length"
openpencil eval design.fig -c "figma.currentPage.selection.forEach(n => n.opacity = 0.5)" -w

Control the running app

When the desktop app is running, omit the file argument — the CLI connects via RPC and operates on the live canvas. Useful for automation scripts, CI pipelines, or AI agents that need to interact with the editor:

openpencil tree                               # Inspect the live document
openpencil export -f png                      # Screenshot the current canvas
openpencil eval -c "figma.currentPage.name"   # Query the editor

All commands support --json for machine-readable output.

AI & MCP

Built-in chat

Press ⌘J to open the AI assistant. It has 100+ tools that can create shapes, set fills and strokes, manage auto-layout, work with components and variables, run boolean operations, analyze design tokens, and export assets. Bring your own API key for OpenRouter, Anthropic, OpenAI, Google AI, Z.ai, MiniMax, or compatible endpoints. No backend, no account.

Not every provider works in the browser, and not every model streams tool calls correctly. See BYOK provider & model compatibility for measured results — contributions welcome.

Coding agents (desktop)

Use Claude Code, Codex, or Gemini CLI directly in the chat panel. The agent connects to the editor's MCP server and uses all 100+ design tools. Requires the desktop app and the agent CLI installed locally.

Pi is also available as an optional AI SDK Harness provider. Install its companion CLI with npm install -g @open-pencil/harness, then add a Pi model profile in Settings → AI & agents. The companion is installed separately so OpenPencil does not bundle a JavaScript runtime for users who do not enable Harness providers.

Setup (Claude Code):

1. Install the ACP adapter: npm install -g @agentclientprotocol/claude-agent-acp
2. Add MCP permission to ~/.claude/settings.json:

   {
     "permissions": {
       "allow": ["mcp__open-pencil__*"]
     }
   }
   

3. Open the desktop app → CtrlJ → select Claude Code from the provider dropdown

MCP server

Connect Claude Code, Cursor, Windsurf, or any MCP client to inspect, modify, and export design documents headlessly. 100+ tools. Full docs →

Stdio (Claude Code, Cursor, Windsurf):

npm install -g @open-pencil/mcp
claude mcp add --scope user open-pencil -- openpencil-mcp

For other MCP clients:

{
  "mcpServers": {
    "open-pencil": {
      "command": "openpencil-mcp"
    }
  }
}

HTTP (scripts, CI):

openpencil-mcp-http   # Unix socket on macOS/Linux + http://127.0.0.1:7600/mcp

Local clients discover the private Unix socket automatically and fall back to localhost TCP. Set PORT=0 to disable TCP on macOS/Linux.

File access: Set OPENPENCIL_MCP_ROOT to scope file operations (open_file, new_document, export path param) to a directory. Defaults to the current working directory.

AI agent skill

Teach your AI coding agent to use OpenPencil — inspect designs, export assets, analyze tokens, modify .fig files:

npx skills add open-pencil/open-pencil

Works with Claude Code, Cursor, Windsurf, Codex, and any agent that supports skills.

For documentation-aware agents, the docs site publishes llms.txt, llms-full.txt, and per-page Markdown files generated from the VitePress docs.

Collaboration

Share a link to co-edit in real time. No server, no account — peers connect directly via WebRTC.

1. Click the share button in the top-right panel
2. Share the generated link (app.openpencil.dev/share/<room-id>)
3. Collaborators see your cursor, selection, and edits in real time
4. Click a peer's avatar to follow their viewport

Why

Figma is a closed platform that actively fights programmatic access. Their MCP server is read-only. figma-use added full read/write automation via CDP — then Figma 126 killed CDP. Your design files are in a proprietary binary format that only their software can fully read. Your workflows break when they decide to ship a point release.

OpenPencil is the alternative: open source (MIT), reads .fig files natively, every operation is scriptable, and your data never leaves your machine.

See the roadmap for product direction and current Figma compatibility gaps.

Contributing

Setup

bun install
bun run dev:portless  # Web editor at https://open-pencil.localhost
bun run dev           # Direct Vite server at http://localhost:1420
bun run tauri dev     # Desktop app (requires Rust)

The first Portless run creates and trusts a local HTTPS certificate. Linked Git worktrees automatically receive branch-prefixed URLs such as https://fix-ui.open-pencil.localhost, so concurrent development servers do not compete for port 1420. Their development MCP bridges are exposed through matching sibling URLs such as https://fix-ui.mcp.open-pencil.localhost, with isolated TCP ports and runtime socket files. Run bunx portless doctor if local routing or certificate trust fails.

Alternatively, open the repository in any Dev Container-compatible tool. The container pins Bun, installs the workspace dependencies, and forwards the direct web editor on port 1420. Start it with bun run dev after the container is ready.

The Dev Container supports the web editor, packages, CLI, and automated checks. Native Tauri development still requires the host setup described below because desktop windows and platform WebView dependencies are not provided in the container.

Quality gates

Command	Description
bun run check	Lint + typecheck
bun run test	E2E visual regression
bun run test:unit	Unit tests
bun run format	Code formatting

Project structure

packages/
  scene-graph/    @open-pencil/scene-graph — nodes, primitives, hit testing, copy/snap/undo
  pen/            @open-pencil/pen — Pencil document format helpers
  kiwi/           @open-pencil/kiwi — Kiwi runtime and low-level .fig container parsing
  fig/            @open-pencil/fig — .fig archives, SceneGraph conversion, instances, metadata
  core/           @open-pencil/core — editor engine, renderer, layout, tools, RPC, document I/O
  dom-css/        @open-pencil/dom-css — HTML/CSS/Tailwind to editable design documents
  vue/            @open-pencil/vue — headless Vue SDK
  cli/            @open-pencil/cli — headless CLI
  mcp/            @open-pencil/mcp — MCP server (stdio + HTTP)
  docs/           Documentation site (openpencil.dev)
src/              Vue app (editor shell, AI, collaboration, document I/O)
desktop/          Tauri v2 desktop app (Rust + config)
tests/            E2E, visual, engine, and integration tests

Tech stack

Layer	Tech
Rendering	Skia (CanvasKit WASM)
Layout	Yoga WASM (flex + grid via fork)
UI	Vue 3, Reka UI, Tailwind CSS 4
File format	Kiwi binary + Zstd + ZIP
Collaboration	Trystero (WebRTC P2P) + Yjs (CRDT)
Desktop	Tauri v2
AI/MCP	Multi-provider (Anthropic, OpenAI, Google AI, OpenRouter), MCP SDK, Hono

Desktop builds

Requires Rust and platform-specific prerequisites (Tauri v2 guide).

bun run tauri build

Acknowledgments

Thanks to @sld0Ant (Anton Soldatov) for creating and maintaining the documentation site.

License

OpenPencil is licensed under the MIT License.

Copyright (c) 2026 Danila Poyarkov and OpenPencil contributors.