convertToExcalidrawElements silently dropped label.text on reload.
Added expandLabelsToNative() that pre-expands server-format labels
into native Excalidraw bound text elements before the scene is set,
ensuring labels survive DB round-trips.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Update lastSyncedElementsRef on every WS-applied scene change so
auto-sync does not revert MCP writes back to stale browser state
- Fix labeled container updates (rect/ellipse/diamond/arrow) to use
convertToExcalidrawElements with bound-text ID transplant, preventing
text clipping and empty labels after update
- Fix standalone text element updates to write into text/originalText
so Excalidraw renders the new value immediately
- Fix convertTextToLabel to handle arrows and empty string text values
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Adds postinstall script to rebuild better-sqlite3 native bindings for
the current Node.js version. Fixes ERR_DLOPEN_FAILED when installing
via npx on a different Node version than was used to publish.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- 9 backend security test files (auth, headers, rate-limit, middleware
order, smoke, validation, WS auth, integration bootstrap)
- 1 e2e test (clear-preference)
- SECURITY.md policy doc
These files powered the 369-test suite but were never committed.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- actions/upload-artifact: replace stale v3 SHA (pinned as v4.6.2) with @v4
- docker/metadata-action: replace non-existent SHA 902fa8ec7 with @v5
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Deletions made in the UI were silently lost on page reload because the sync
baseline (lastSyncedElementsRef) was never populated after initial load,
making the delta algorithm unable to detect removed elements. Additionally,
import_scene and restore_snapshot used a non-atomic clear+create pattern
that could permanently lose all canvas data if the batch create failed
after clearing, and duplicate_elements copied stale binding references
pointing to original element IDs instead of remapped duplicates.
🔧 Sync baseline restoration:
- Populate deletion-detection baseline on every server-to-client data path
(page load, delta resync, hello handshake, initial elements broadcast)
- Establish sync version and hash baselines to prevent phantom re-syncs
🛡️ Data-loss prevention:
- Backup current scene before destructive clear in replace-mode operations
- Atomic restore from backup when subsequent batch create fails
- Remap all binding references (start/end IDs, boundElements, containerId)
to new IDs during element duplication
✅ Comprehensive test coverage (154 new tests, 344 total):
- Delta sync flows including deletion persistence and bidirectional sync
- Multi-tenant element/sync/WebSocket isolation
- Arrow binding resolution across all shape types and edge cases
- MCP tool integration covering backup-restore and binding remapping
- Input validation and security boundary testing
- Frontend sync algorithm unit tests reproducing the exact regression
🎯 Eliminates the most critical data-integrity risks: deletions now
persist reliably, destructive operations are rollback-safe, and the
full test suite provides regression coverage for every sync path.
Fix 6 bugs discovered during MCP tool usage:
1. syncToCanvas error handling: Distinguish network errors (return null)
from API errors (re-throw with actual message). Fixes misleading
"HTTP server unavailable" on batch_create_elements.
2. USER_PREFS fallbacks: create_element and batch_create_elements now
apply fontFamily/roughness/fontSize/strokeWidth from preferences.json
when not explicitly provided by the caller.
3. Hello handshake: Frontend sends `hello` on tenant_switched and handles
`hello_ack`. Server resolves projectId from tenantId when absent.
Fixes WS connections being registered under wrong scope.
4. Serialized broadcasts: Add serializedBroadcastWithAck() that queues
broadcasts per tenant/project scope. Prevents race condition where
parallel MCP create_element calls produce overlapping WS messages
that clobber each other in the frontend.
5. Viewport screenshot: get_canvas_screenshot passes captureViewport=true,
frontend captures DOM canvas via toDataURL() instead of exportToBlob()
which always rendered the full scene bounding box.
6. Viewport animate:false: set_viewport uses animate:false for instant
positioning, preventing mid-animation screenshot captures.
Tests: 14 new tests (8 API, 6 WS) + 9 E2E specs covering all fixes.
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Implement a complete sync architecture overhaul (12 tasks) replacing the flat
WebSocket broadcast with scoped, acknowledged delivery:
**Backend (server.ts, db.ts, types.ts, index.ts):**
- Scoped connection registry: Map<tenant, Map<project, Set<ClientConnection>>>
- Hello handshake: WS clients identify tenant/project, server responds with scoped elements
- broadcastToScope() replaces global broadcast for element mutations
- broadcastWithAck() waits for browser ACK before returning syncedToCanvas status
- sync_version: monotonic counter per project, stamped on every mutation
- Delta sync v2: POST /api/elements/sync/v2 for incremental sync with version tracking
- GET /api/sync/version endpoint
- Honest syncedToCanvas + canvasStatus in all mutation responses
- Fixed silent try/catch in tenant switch verification
**Frontend (App.tsx):**
- ACK sending after every updateScene() with element verification
- Delta sync v2 integration in syncToBackend()
- Gap detection: triggers resync when sync_version gaps are detected
- lastSyncVersion tracking via refs + localStorage persistence
**Tests (40 new tests, 168 total):**
- db.test.ts: +11 tests for sync_version CRUD, scoping, getChangesSince
- ws.test.ts: +8 tests for hello handshake, scoped broadcast, ACK model
- api.test.ts: +10 tests for sync/v2, sync/version, canvasStatus responses
- helpers.test.ts: +11 tests for isImageElement, normalizeImageElement, restoreBindings
- canvas.spec.ts: +8 e2e tests including full ACK pipeline verification
- Fixed stale tenant state bug in api.test.ts beforeEach
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Font family IDs were duplicated across 5 files (types.ts, setup.ts,
index.ts, SKILL.md, preferences.example.json) with inconsistent
mappings — Comic Shanns was 4 in some places but actually 8 in
Excalidraw source. This caused wrong fonts to render on canvas.
Fix: create src/font-families.json as the canonical font data, import
it in types.ts, and derive all other references from it. Static docs
now point to the JSON file instead of duplicating the mapping.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Users are now prompted to choose their preferred font family and roughness
style during both `setup` and `update`. Preferences are saved to
~/.claude/skills/excalidraw-skill/preferences.json, which the MCP server
already reads at startup via loadPreferences(). This ensures third-party
users who install via npx get preferences configured before first use.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add a preference system that lets users configure default font, roughness,
fontSize, and strokeWidth — with three scopes (session/folder/global).
Skill layer (Step 1 in SKILL.md):
- Reads .claude/excalidraw-preferences.json (folder) then
~/.claude/skills/excalidraw-skill/preferences.json (global)
- If neither exists, prompts user interactively on first use
- Session-only scope keeps preferences in-memory without saving
Server layer (index.ts):
- loadPreferences() reads the same files at startup
- Replaces hardcoded fontFamily ?? 1 with USER_PREFS.fontFamily
- Folder-level preferences override global; user values override both
Also ships preferences.example.json as a template (preferences.json
is gitignored).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The excalidraw-skill was not being auto-invoked when users prompted
Claude to draw diagrams, despite being installed. Claude would call
Excalidraw MCP tools directly, bypassing the skill's critical sizing
formulas and verification workflow — producing broken diagrams with
invisible arrows, truncated text, and overlapping elements. The root
cause is that Claude Code's skill system is advisory: when direct MCP
tools or built-in Bash instructions are available, Claude skips skill
consultation entirely.
🔧 Skill description rewrite:
- Lead with "MANDATORY prerequisite" to assert priority over raw MCP tools
- Add user-intent trigger keywords (draw, visualize, sketch, diagram)
- Name specific consequences of skipping the skill
- List common diagram types for semantic matching
🏗️ Setup/update auto-directives:
- Write marked CLAUDE.md sections during skill install and update
- Write Cursor .mdc rules with alwaysApply for Cursor users
- Use HTML comment markers for idempotent append-or-replace on updates
- Non-fatal directive writing — skill installs even if directive fails
🎯 Two-layer defense ensures reliable skill triggering: the description
catches semantic matching, while the CLAUDE.md directive provides an
authoritative instruction that Claude cannot deprioritize in favor of
raw tool access.
The `claude mcp add` command fails with "already exists" when the server
is already registered. Additionally, configs written by setup/update used
a bare package name without @latest, causing npx to serve stale cached
versions indefinitely.
🔧 CLI registration:
- Remove existing entry before re-adding for Claude Code (both setup and update flows)
- Silently ignore removal errors when entry doesn't exist yet
⚡ Auto-update via @latest:
- Write @latest suffix in JSON configs, CLI commands, and manual instructions
- Smart detection in update flow: skip if config already has @latest,
offer migration if pinned, offer creation if missing
🎯 Users who run setup or update now get configs that always fetch the
newest version on MCP client restart, eliminating the stale-cache problem
without requiring manual npm cache clearing.
Existing users who update the npm package often forget to update the agent
skill files, leaving their AI agent working with stale workflow guidance
(sizing rules, color palettes, anti-patterns). The new `update` subcommand
solves this by detecting and updating all skill installations automatically.
🔧 Interactive update wizard:
- Scan all agent directories (Cursor, Claude Code, Codex CLI) across global
and local scopes for existing excalidraw-skill installations
- Batch-update found installations with a single Y/n confirmation
- Offer to install the skill for detected agents that don't have it yet
- Optionally re-apply MCP config (defaults to no for non-breaking updates)
📝 Documentation:
- Rewrite README "Updating" section to lead with the interactive command
- Add collapsible example session showing the update flow
- Retain manual update methods as fallback for each installation path
🎯 Ensures skill files stay in sync with MCP tools across releases,
preventing the subtle drift where new tool capabilities exist but the
agent's workflow guidance doesn't reference them.
Excalidraw skill lacked guidance on building complex visual shapes from
basic primitives, leading to flat single-shape attempts for structures
like roofs and walls. Adds coordinate geometry foundations so agents
produce realistic illustrative diagrams alongside technical ones.
🔧 Core geometric principles:
- Compose complex shapes from many small primitives with tessellation
- Eliminate gaps using interleaving offset rows at half-width spacing
- Parametric positioning formulas for centering, circular, triangular, and isometric layouts
- Scale primitive dimensions proportionally to container size
📦 Technique catalog reference:
- Illustrative elements (roofs, walls, clouds, trees, fences, windows)
- Sugiyama-inspired flow diagram layout algorithm
- Zone-grid architecture diagram layout
- Isometric 2.5D projection formulas
- Repeating pattern formulas with brick-pattern offsets
🧹 Consolidate tenants/projects/search into single compact table to
keep SKILL.md under the 500-line skill guideline
The postinstall ran prebuild-install and node-gyp in the wrong context
(our package instead of better-sqlite3), always failing with "binding.gyp
not found". better-sqlite3 handles its own native compilation via its
install script — the package-level postinstall was redundant.
Co-authored-by: sanjibdevnathlabs <devnath.sanjib@gmail.com>
- Use fs.realpathSync for entry point detection to fix npx symlink failures
- Add --help and --version CLI flags with explicit process.exit(0)
- Add TTY detection in setup wizard to prevent non-interactive hangs
- Wrap JSON.parse in mergeJsonConfig with try/catch for malformed configs
- Update engines.node to >=20.0.0 to match better-sqlite3 requirements
- Update Dockerfiles from node:18-slim to node:20-slim
- Remove error-swallowing || true from postinstall script
- Replace deprecated windows-build-tools with VS Build Tools link
The release workflow pushes version bump commits directly to main, but
branch protection rules block the default GITHUB_TOKEN from bypassing
required status checks and PR requirements. Using a dedicated GitHub
App (sanjibdevnathlabs-release-bot) generates installation tokens that
are permitted through the ruleset bypass list, and keeps the bot
identity on release commits instead of a personal account.
The Docker Build workflow previously used a paths filter, causing it to
not trigger at all for non-Docker PRs — leaving the required
github/docker-build-check status permanently pending. Now the workflow
always triggers but checks for Docker-related file changes first,
skipping builds when unnecessary while still reporting the status check
as passed.
Establish comprehensive quality infrastructure for a project that previously
had zero tests, enabling confident refactoring and community contributions
with automated guardrails. Port upstream enhancements for font normalization,
image element support, and arrow binding preservation.
🏗️ Testing infrastructure:
- Unit tests for SQLite persistence layer and element validation helpers
- Integration tests for REST API, WebSocket broadcast, and arrow binding
- E2E tests with Playwright for canvas rendering and real-time sync
- Vitest + Playwright configuration with proper isolation
👷 CI/CD pipeline:
- Auto-versioning from conventional commits on push to main
- Auto-publish to NPM and Docker Hub on GitHub release
- Matrix testing across Node 18/20/22 with pinned dependencies
- Docker health check with diagnostic logging on failure
- Preserve rollup status checks for branch protection gates
📦 Developer experience:
- Interactive setup wizard for first-time configuration
- Canvas clear confirmation and scene description tools
- Frontend helpers extracted for testability
🔧 Upstream feature ports:
- Font family normalization (string names to numeric IDs)
- Image element support with file management API
- Arrow binding preservation through server round-trips
- Vite config fix for font subsetting worker chunk names
- Idempotent database initialization for standalone Docker mode
🐛 Docker fixes:
- Set EXCALIDRAW_DB_PATH in both Dockerfiles to writable /app/data/
- Make initDb() idempotent and closeDb() reset-safe for test isolation
🎯 Provides the safety net needed for rapid iteration — every PR is
validated across 120 test cases before merge, and releases are fully
automated from commit to published package.
Co-authored-by: sanjibdevnathlabs <devnath.sanjib@gmail.com>
The npm-publish workflow needs write access to upload tarballs to
GitHub releases. Read-only caused "Resource not accessible by integration".
Co-authored-by: Cursor <cursoragent@cursor.com>
* 🔧 chore: reset version to 1.0.0 for independent NPM package
This is a new scoped package (@sanjibdevnath/mcp-excalidraw-local),
so versioning starts fresh at 1.0.0 rather than continuing upstream's
numbering.
Co-authored-by: Cursor <cursoragent@cursor.com>
* chore: trigger PR
Co-authored-by: Cursor <cursoragent@cursor.com>
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
Dogfood our own tool — the architecture diagram is now drawn in
Excalidraw and exported as PNG, replacing the ASCII art block.
Co-authored-by: Cursor <cursoragent@cursor.com>
