Figma's parentIndex.position strings represent z-order, not visual
layout order. In auto-layout frames, children must be sorted by their
x (horizontal) or y (vertical) coordinate to match the visual order.
Fixes scrambled calendar dates and other auto-layout child ordering.
Each instance in Figma's clipboard carries symbolData.symbolOverrides
with per-child property changes (text, fills, visibility). The override
targets a child via guidPath which matches the overrideKey field on the
component's child node.
After cloning component children into instances, overrides are resolved
through the overrideKey→figmaId→internalId→clonedChild chain and applied
to the instance's children.
FRAME nodes with VARIANT componentPropDefs are now promoted to
COMPONENT_SET on import (both .fig files and clipboard paste).
Clipboard paste now respects the Internal Only Canvas: components
on it are used to populate instance children but are not pasted
as visible nodes, matching Figma's behavior.
When pasting from Figma, INSTANCE nodes reference their component via
symbolData.symbolID but have no children in the clipboard. If the
component was also pasted, clone its children into the instance using
the componentId mapping (cloneChildrenWithMapping).
Figma vectors use a normalized coordinate space (vectorData.normalizedSize)
that can differ from the node's actual size. The decoded path vertices and
tangents must be scaled to fit the node bounds, otherwise vectors render
at the wrong size (e.g. 16×16 path in a 12×12 node).
Extract shared kiwi→SceneNode conversion into kiwi-convert.ts, used by
both fig-import.ts (.fig file opening) and clipboard.ts (Figma paste).
The clipboard import was a simplified copy that diverged over time:
- Missing SYMBOL→COMPONENT type mapping (Figma's kiwi name for components)
- textAutoResize hardcoded to NONE instead of using clipboard value
- No gradient/image fill support, no effects, no style runs
- Missing flipX/flipY, constraints, text decorations, arc data, etc.
- Percent letter spacing not converted to pixels
Now both paths use nodeChangeToProps() with identical conversion logic.
Decode internalOnly from kiwi field 142, filter internal pages
in SceneGraph.getPages() by default (pass true to include them).
Pages with names of only dashes/asterisks/spaces and no children
render as horizontal divider lines.
Panel resizing now creates a new surface from the existing
GrDirectContext instead of destroying and recreating the entire
renderer. All caches (paints, fonts, node pictures, scene picture,
filters) are preserved across resizes.
* Replace saveLayer with MaskFilter for drop shadows
Drop shadows now draw directly with MaskFilter.MakeBlur instead of
saveLayer + ImageFilter. Eliminates per-shadow offscreen buffer
allocation, significantly improving render throughput for scenes
with many shadowed nodes.
Text shadows still use saveLayer (MaskFilter can't be applied to
text glyphs individually).
Benchmark: 50 shadow nodes render at 1.55ms/frame (scene change).
* Cache per-node SkPictures for effect rendering
Nodes with visible effects (shadows, blurs) are recorded into
individual SkPictures and replayed on subsequent scene redraws.
Only the modified node re-records; unchanged shadow nodes replay
from cache.
- MaskFilter.MakeBlur for drop shadows instead of saveLayer
- Reusable effectLayerPaint and cached ImageFilters/MaskFilters
- invalidateNodePicture() called on node update
- invalidateAllPictures() on font load; invalidateScenePicture()
preserves node cache for scene-only invalidation
- Shadow benchmark: 50 nodes at 1.47ms/frame (scene change)
* Update changelog with shadow rendering performance improvements
Eliminates per-frame WASM allocations for shadows and blurs:
- Reusable effectLayerPaint instead of new Paint() per effect
- ImageFilter cache keyed by params (drop shadow, blur, decal blur)
- Zero allocations during SkPicture recording for cached effects
* Fix shadow rendering, implement blur effects, add spread support
- Fix drop shadow rendering order: draw shadows before fills so they
appear behind opaque shapes instead of on top (#23.1)
- Implement shadow spread: expand/contract shadow shape by spread value
for both drop shadow and inner shadow (#23.2)
- Implement background blur: clip to node shape and apply blur filter
to content behind the node (#23.3)
- Fix text shadows: use saveLayer with ColorFilter.MakeBlend to apply
shadows to text glyphs instead of the bounding box (#23.4)
- Implement layer blur with actual saveLayer + blur filter instead of
the no-op comment (#23.3)
- Implement foreground blur: clip to node shape and blur content in
front (#23.7)
- Add inner shadow spread support with rounded rect handling
- Add makeRRectWithSpread and makeRRectWithOffset helpers
Closes#23
* Add Effects section to demo showcasing shadow and blur fixes
Drop shadows (subtle, medium, heavy, with spread), inner shadows
(inset, with spread, on ellipse, combined), text shadows on glyphs,
layer blur, and glassmorphism background blur card.
* Fix drop shadow, layer blur, and background blur rendering
Drop shadow: use MakeDropShadowOnly to produce shadow-only output
that doesn't bleed through opaque fills. The sigma is radius/2 to
match Figma's blur convention.
Layer blur: wrap entire node content in a saveLayer with blur filter
at the renderNode level instead of trying to blur after-the-fact
in renderEffects.
Background/foreground blur: extract clipNodeShape helper to reduce
duplication in blur effect rendering.
Update tests to match new rendering approach.
* Add visual regression tests for effects rendering
9 snapshot tests covering drop shadow, inner shadow, spread,
ellipse shapes, text glyphs, layer blur, and combined effects.
* Make Effects demo section background white
* Clean up renderer code and tests
- Extract applyClippedBlur to deduplicate background/foreground blur
- Remove comments that repeat what code does
- Format with oxfmt
* Add undo/redo support for effect property changes
ScrubInput controls (offset, radius, spread, opacity) now use
scrubEffect for live preview and commitEffect on release to create
a single undo entry per interaction.
Cherry-picked the 2 unique tools from PR #11:
- set_text_properties: alignment, auto-resize, decoration
- set_layout_child: sizing, grow, align_self, positioning
The other 10 tools in the PR were already covered by the
port-tools merge (73 → 75 tools).
13 tests using InMemoryTransport from @modelcontextprotocol/sdk:
- Tool registration (all 76 tools listed with descriptions)
- new_document, open_file, save_file lifecycle
- Tool execution through MCP protocol (create, fill, delete, query)
- Error paths: no document loaded, nonexistent node, invalid file path
- Full workflow: new → create → nest → query → delete
Fix @open-pencil/core dep in MCP package: workspace:* for local dev
(pnpm publish resolves this to the actual version at publish time).
On copy: serialize each text node's paragraph into an SkPicture and
embed the bytes (base64) in the OpenPencil clipboard payload.
On paste: if the font isn't available, render from the cached SkPicture
instead of the broken fallback. Text looks pixel-perfect without the
font installed — same behavior as Figma.
Missing font indicator:
- Orange ⚠ badge on canvas over text nodes with unavailable fonts
- Amber warning banner in the Typography panel listing missing families
- isFontLoaded() check in @open-pencil/core for querying font status
Replace 8 inline { r: 0, g: 0, b: 0, a: 1 } literals across core
with shared constants from constants.ts. Also reuse CANVAS_BG_COLOR
in fig-export.ts instead of duplicating the value.
- Use Int32Array instead of Uint32Array for Kiwi pasteID (signed int field)
- Error toasts: don't auto-dismiss, show copy button, text is selectable
- Add clipboard roundtrip tests (encode → decode → verify)
- letterSpacing is an object {value, units} not a number — was NaN,
which silently breaks CanvasKit paragraph layout
- Add mapLetterSpacing() to handle PIXELS and PERCENT units
- Force textAutoResize to NONE (Yoga has no text measurement callback)
- Record undo entry for Figma paste with full subtree snapshots
- Add stackChildAlignSelf, fontWeight to NodeChange interface
Figma clipboard data includes VARIABLE_SET, VARIABLE, and other non-visual
types that were falling through to the default RECTANGLE case in mapNodeType.
These now join DOCUMENT and CANVAS in the skipTypes set.
The first render records an SkPicture before fonts are loaded (fallback
drawText path). When fonts finish loading asynchronously, the cached
picture was never invalidated — subsequent renders replayed stale picture
without paragraph text. Hovering a node forced a full re-render (volatile
overlay path), making text reappear. Un-hovering replayed the stale cache.
Fix: call invalidateScenePicture() at the end of loadFonts() so the next
render records a fresh picture with proper paragraph text.
29 tools: open_file, save_file, new_document + all 26 from @open-pencil/core.
Stdio for MCP clients (Claude Code, Cursor), HTTP (Hono + Streamable HTTP
with sessions) for scripts, browser extensions, CI.
Runs on both Bun and Node.js (via tsx).
Penpot 2.x ships with Rust/Skia WASM rendering enabled by default.
Update architecture diagram, rendering pipeline, and file format sections
to reflect the current state instead of the old SVG-based renderer.
- Cloudflare Worker + Durable Object server (packages/collab/)
- One DO per room with WebSocket Hibernation for near-zero idle cost
- Yjs sync protocol for document state (CRDT merge)
- Yjs awareness protocol for cursors, selection, presence
- Persists doc state to DO storage
- CORS headers for cross-origin WebSocket upgrade
- Client composable (src/composables/use-collab.ts)
- Y.Doc ↔ SceneGraph bidirectional binding
- Awareness: broadcasts cursor position, selection, name, color
- Auto-join from /share/:roomId URL
- Suppresses echo loops between Yjs and graph updates
- CollabPanel component (src/components/CollabPanel.vue)
- Share button generates room link, copies to clipboard
- Join flow: paste link or room ID
- Avatar stack with colored initials and tooltips
- Name input persisted to localStorage
- Remote cursor rendering in CanvasKit
- Colored arrow cursor + name label pill per remote peer
- Remote selection outlines in peer's color
- Only shows cursors on the same page
- SPA redirect for /share/* routes (Cloudflare Pages _redirects)
expanded: was mapped from frameMaskDisabled (unrelated Kiwi field),
now defaults to true. strokeMiterLimit: was mapped from speculative
strokeMiterAngle field that doesn't exist in Kiwi codec, now uses
hardcoded default 4.
- Replace _id/_graph/_api with Symbol-keyed properties so eval scripts
can't access SceneGraph internals directly
- fills/strokes/effects return frozen structuredClone copies
- Fix currentPage getter to define selection only once per proxy
- Fix layoutSizingHorizontal/Vertical: respect parent layout direction
even when the node itself is an auto-layout frame
- Add 6 new tests: layoutSizing, frozen arrays, internals not exposed
expanded: was mapped from frameMaskDisabled (unrelated Kiwi field),
now defaults to true. strokeMiterLimit: was mapped from speculative
strokeMiterAngle field that doesn't exist in Kiwi codec, now uses
hardcoded default 4.
- Replace _id/_graph/_api with Symbol-keyed properties so eval scripts
can't access SceneGraph internals directly
- fills/strokes/effects return frozen structuredClone copies
- Fix currentPage getter to define selection only once per proxy
- Fix layoutSizingHorizontal/Vertical: respect parent layout direction
even when the node itself is an auto-layout frame
- Add 6 new tests: layoutSizing, frozen arrays, internals not exposed
- sceneNodeToJsx() in @open-pencil/core: converts SceneNode subtree to JSX
- CodePanel.vue with Prism.js highlighting, line numbers, copy button
- Third tab in properties panel: Design | Code | AI
- 14 new tests covering shapes, text, layout, effects, multi-selection
- Track click count with timing/proximity in onMouseDown
- Double-click in text: select word (was only handled in dblclick)
- Triple-click in text: select all
- Double-click to enter editing: select word at click position
- Add selectLine/selectLineAt to TextEditor
Replace src/kiwi/ copies with re-exports from @open-pencil/core.
Extract shared sceneNodeToKiwi into kiwi-serialize.ts, used by
both fig-export.ts and clipboard.ts.
jscpd: 15.6% → 1.7% duplication (35 → 18 clones, -4459 lines)
- analyze colors: color palette usage with clustering
- analyze typography: font/size/weight distribution
- analyze spacing: gap/padding values with grid check
- analyze clusters: repeated patterns (potential components)
- node: detailed properties by ID
- pages: list pages with node counts
- variables: list design variables and collections
StyleRun model: array of {start, length, style} with
CharacterStyleOverride (fontWeight, italic, textDecoration, etc.).
Renderer uses ParagraphBuilder.pushStyle/pop for mixed-style text.
TextEditor delegates paragraph building to renderer.buildParagraph().
⌘B/I/U apply to selection range when text is selected, or toggle
whole-node style when no selection. Style runs adjust on
insert/delete to preserve formatting across edits.
Fig roundtrip: import characterStyleIDs + styleOverrideTable from
TextData, export back with deduped style table.
B/I/U/S toggle buttons in TypographySection.
TextEditor class in core — cursor positioning, selection, word
boundaries, line navigation via CanvasKit Paragraph API
(getGlyphPositionAtCoordinate, getRectsForRange).
Renderer draws selection highlights and blinking caret directly
on the canvas. Blue outline around text node during editing.
Phantom textarea captures input, IME composition, clipboard,
arrow keys, Home/End, word/line movement (⌥/⌘ modifiers).
Replaces the old visible <textarea> overlay.
Rust: list_system_fonts and load_system_font Tauri commands using
font-kit crate for cross-platform system font access.
TS: font shim detects Tauri runtime and uses invoke() for font
listing/loading instead of queryLocalFonts (unavailable in WKWebView).
Commands:
- open-pencil info <file> — document stats, node types, fonts
- open-pencil tree <file> — visual node tree with agentfmt
- open-pencil find <file> — search by name/type with rich output
- open-pencil export <file> — render to PNG/JPG/WEBP at any scale
Uses canvaskit-wasm/full for headless CPU rasterization.
All commands support --json for machine-readable output.
agentfmt provides consistent colored terminal formatting.
Move engine (scene-graph, renderer, layout, codec, kiwi) into
packages/core/ as a Bun workspace package. The app's src/engine/
files become thin re-export shims so all existing imports still work.
- Bun workspace with packages/core
- Zero DOM dependencies in core
- Headless-friendly CanvasKit init (locateFile option)
- Constants split: engine constants in core, UI constants in app
- All 76 unit tests pass, lint + typecheck clean, vite build works
