About Skedra Whiteboard
Skedra Whiteboard is a self-hosted collaborative infinite canvas for teams and AI agents to create and edit boards. It combines MCP access with Kanban, Gantt charts, diagrams, and real-time collaboration.
Key features
- 01Infinite canvas with drawing elements
- 02Flowcharts, Kanban, and Gantt charts
- 03Real-time collaboration and permissions
- 04Share links and read-only embeds
- 05MCP server with 23 tools
- 06Self-hosting and React SDK
Use cases
- 01Create diagrams collaboratively with a team
- 02Plan tasks on Kanban boards and Gantt timelines
- 03Have AI agents edit canvas content
- 04Present read-only boards through share links
MCP support
The built-in MCP server connects to a self-hosted Skedra instance and exposes 23 tools for editable agent workflows. Clients can connect using OAuth, API keys, or local stdio configuration.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://skedra.xyz/api/mcpclaude mcp add --transport http moonriddim-skedra https://skedra.xyz/api/mcp{"mcpServers":{"moonriddim-skedra":{"type":"http","url":"https://skedra.xyz/api/mcp"}}}Open in Cursor{"servers":{"moonriddim-skedra":{"type":"http","url":"https://skedra.xyz/api/mcp"}}}Open in VS Code{"mcpServers":{"moonriddim-skedra":{"type":"http","url":"https://skedra.xyz/api/mcp"}}}{"mcpServers":{"moonriddim-skedra":{"serverUrl":"https://skedra.xyz/api/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project README
The open-source whiteboard for people and AI agents.
Self-hosted. Collaborative. Agent-editable. AGPL-3.0.
Website · Documentation · Self-host · MCP · React SDK · Discord
Skedra is an infinite canvas for sketches, diagrams, Kanban boards and Gantt timelines. Work with your team in real time, or connect AI agents through the built-in MCP server. Run the complete application on your own infrastructure. All features are available without a Skedra subscription, license server or central Skedra account.
skedra.xyz provides documentation and project information. The application, API, MCP server, library catalog, database, SDK and deployment files live together in this repository. There is no separate commercial edition.

Self-host Skedra
Run the all-in-one Docker image:
docker run -d \
--name skedra \
-p 3000:80 \
-v skedra_data:/data \
ghcr.io/moonriddim/skedra-community-standalone:latest
Open http://localhost:3000. PostgreSQL, files and instance
secrets persist in the skedra_data volume. The first registered user becomes the
instance administrator. Further registrations require an invitation by default.
See SELFHOST.md for Docker Compose, production URLs, storage, optional LiveKit calls, updates and backups. Images for this transition become available after the changes are merged and the release workflow succeeds.
Features
- Infinite canvas with text, shapes, arrows, drawing, images and frames.
- Native flowcharts, sequence diagrams, Kanban boards, Gantt charts and templates.
- Local browser boards and portable
.skedrafiles. - Accounts, teams, roles, permissions, comments, mentions and board activity.
- Real-time collaboration, encrypted storage and optional end-to-end encryption.
- Share links, presentations and read-only embeds.
- MCP server with 23 tools for structured, editable agent workflows.
- Shape libraries with
.skedralibimport and a self-hosted catalog. - Optional AI with your own provider or local models, and optional LiveKit calls.
- Embeddable React SDK using the same canvas model, renderer and editor.
MCP for AI agents
Connect Codex, Claude, Cursor or another client to your own instance:
https://YOUR-SKEDRA-INSTANCE/api/mcp
Sign in to that instance through OAuth. API-key and local stdio configurations are available in Settings > API Keys & MCP. There is no central hosted MCP endpoint. See docs/MCP.md.
Development
git clone https://github.com/moonriddim/skedra-community.git
cd skedra-community
corepack enable
docker compose -f docker-compose.dev.yml up -d
cp .env.example .env
pnpm install
pnpm db:push
pnpm dev
Open http://localhost:5174.
pnpm dev also watches the shared SDK. Use pnpm sdk:verify for focused SDK and
editor parity checks. pnpm test runs the workspace tests.
Build the independent static website with pnpm docs:build and preview it with
pnpm docs:dev at http://localhost:5176.
It requires no API, database, authentication or external service.
Repository
apps/web,apps/api,apps/mcp,apps/libraries: self-hosted application.apps/docs: static project website and documentation.packages: database, shared types, canvas core, editor, renderer, IO and React SDK.deploy: self-hosting, database migrations and reverse proxies..github/workflows: checks and releases from this repository directly.
Contributions are welcome. Read CONTRIBUTING.md, report bugs in GitHub Issues, or join Discord.
License
Skedra is licensed under AGPL-3.0-only. You may use, modify, redistribute and host it under that license, including commercially. Third-party material retains its own terms. Your original boards and exports remain yours. See LICENSING.md and the transition guide.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.