Get started with Maskin

Maskin is an open-source workspace where your whole team — humans and AI agents — works together with shared memory, context, and process. This guide connects Claude to the hosted Maskin MCP so you can drive a workspace from inside Claude Code or Claude Desktop. Expect to spend about five minutes.

Prerequisites

• Claude Code or Claude Desktop installed.
• An Anthropic API key. Maskin runs Claude Code in sandboxes on your behalf and uses this key to authenticate those runs.

You don't need to clone anything, install dependencies, or run a local server. Maskin is hosted at maskin.sindre.ai.

Step 1 — Sign in and create a workspace

1. Go to maskin.sindre.ai and sign in.
2. Create your first workspace from the prompt. A workspace is an isolated environment with its own objects, actors, and settings — you can create more later.

Once you're in, you'll see the unified pipeline view: Insights → Bets → Tasks. All work — yours and your agents' — lives as objects in this pipeline.

Step 2 — Paste your Anthropic API key

Maskin needs your Anthropic API key to run agent sessions in its sandboxes. Keys are stored encrypted and scoped to the workspace.

1. Open Settings → Integrations (or Settings → LLM keys) from the workspace sidebar.
2. Paste your Anthropic API key and Save. Maskin validates the key with a cheap call to Anthropic before storing it.
3. You can update or remove the key from the same screen at any time. After save, only the last four characters are shown.

Step 3 — Get your Maskin API key and workspace ID

To let Claude talk to your workspace, you need two values from the UI:

1. Open Settings → API keys and click Create key. Copy the ank_… value that appears — this is your Maskin API key. You'll only see it once.
2. Open Settings → Workspace and copy the Workspace ID (UUID). This tells the MCP which workspace to act on.

Keep both values handy for the next step.

Step 4 — Connect Claude to the hosted Maskin MCP

Claude Code

From a terminal, run:

claude mcp add maskin \
  --transport http \
  --url https://maskin.sindre.ai/mcp \
  --header "Authorization: Bearer <YOUR_MASKIN_API_KEY>" \
  --header "X-Workspace-Id: <YOUR_WORKSPACE_ID>"

Replace <YOUR_MASKIN_API_KEY> with the ank_… key from Step 3 and <YOUR_WORKSPACE_ID> with the workspace UUID.

If Claude Code is already running, reload plugins so the new tools show up in the current session:

/reload-plugins

Claude Desktop

Open your Claude Desktop MCP config (Settings → Developer → Edit config) and add the maskin entry:

{
  "mcpServers": {
    "maskin": {
      "type": "http",
      "url": "https://maskin.sindre.ai/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_MASKIN_API_KEY>",
        "X-Workspace-Id": "<YOUR_WORKSPACE_ID>"
      }
    }
  }
}

Save the file and restart Claude Desktop.

Step 5 — Make your first MCP call

Back in Claude Code or Claude Desktop, ask:

Use the maskin MCP to call get_started.

get_started is the entry-point tool. It previews and applies a workspace template — either development (for shipping a product), growth (for running a launch or outreach pipeline), or custom (a short questionnaire that tailors the workspace to whatever you're working on). Once you pick one, Claude seeds the workspace with object types, statuses, custom fields, and a first set of objects, and kicks off the pipeline.

If the tool returns a response without errors, you're connected. From here Claude can create and update insights, bets, and tasks, run triggers, and stream events directly against your hosted workspace.

What's next

• Invite teammates. Workspaces support multiple members — add humans and agents from Settings → Members.
• Wire up integrations. Connect Slack, GitHub, Google Calendar, and more from Settings → Integrations to let agents react to real-world events.
• Explore the 39 MCP tools. get_started is just the start — list objects, create bets, fire triggers, and manage agent sessions all from Claude. Run list_tools to see the full set.

Stuck? Open an issue in sindre-ai/maskin or read the README for deeper background on the data model, agent sessions, and the event stream.