dominodatalab

Domino Data Lab Plugin for Claude Code

A comprehensive Claude Code plugin providing full coverage of the Domino Data Lab platform for AI-assisted development.

Overview

This plugin enables Claude Code to help you with all aspects of Domino Data Lab, including:

• Workspaces: Jupyter, VS Code, RStudio configuration and management
• Jobs: Batch execution, scheduled jobs, and monitoring
• Environments: Custom Docker environments and package management
• Datasets: Data versioning with snapshots and sharing
• NetApp Volumes: Enterprise-grade multi-terabyte storage with near-instant snapshots
• Apps: Deploy React, Streamlit, and Dash applications
• Models: Deploy, monitor, and manage model endpoints
• GenAI: Trace and evaluate AI agents with the Domino SDK
• Distributed Computing: Spark, Ray, and Dask clusters
• And more...

Installation

Prerequisites

• Claude Code CLI v1.0.33 or later (claude --version to check)
• Access to a Domino Data Lab instance
• Domino API key (for API operations when running outside a Domino workspace)
• uv package manager (install guide) — required for the bundled Domino MCP server

---

Option 1: Marketplace Install (Recommended)

This approach registers the plugin through Claude Code's native marketplace system so it persists across sessions.

Step 1: Clone the repository and create a marketplace wrapper

# Clone the plugin
git clone https://github.com/dominodatalab/domino-claude-plugin.git

# Create the marketplace directory structure
mkdir -p ~/.claude/marketplaces/domino/.claude-plugin
mkdir -p ~/.claude/marketplaces/domino/plugins

# Move the plugin into the marketplace
mv domino-claude-plugin ~/.claude/marketplaces/domino/plugins/domino-claude-plugin

Step 2: Create the marketplace manifest

cat > ~/.claude/marketplaces/domino/.claude-plugin/marketplace.json << 'EOF'
{
  "name": "domino-marketplace",
  "owner": {
    "name": "Domino Data Lab",
    "email": "support@dominodatalab.com"
  },
  "plugins": [
    {
      "name": "domino-claude-plugin",
      "description": "Domino Data Lab plugin for Claude Code - workspaces, jobs, environments, datasets, apps, models, and more",
      "version": "1.0.0",
      "source": "./plugins/domino-claude-plugin",
      "category": "development"
    }
  ]
}
EOF

Step 3: Register the marketplace and install the plugin

Launch Claude Code and run:

/plugin marketplace add /home/<your-username>/.claude/marketplaces/domino
/plugin install domino-claude-plugin@domino-marketplace

> Note: Replace <your-username> with your actual username, or use the full absolute path (e.g., /home/ubuntu/.claude/marketplaces/domino). The ~ shorthand may not expand correctly.

Step 4: Restart Claude Code

/exit
claude

The plugin should appear in your loaded plugins on startup. Verify with:

/plugin

Navigate to the Installed tab to confirm domino-claude-plugin is listed.

---

Option 2: Direct Plugin Directory (Development / Quick Start)

Use the --plugin-dir flag to load the plugin directly. This is ideal for development, testing, or quick evaluation.

# Clone the plugin
git clone https://github.com/dominodatalab/domino-claude-plugin.git

# Ensure the plugin manifest exists
mkdir -p domino-claude-plugin/.claude-plugin
cat > domino-claude-plugin/.claude-plugin/plugin.json << 'EOF'
{
  "name": "domino-claude-plugin",
  "description": "Domino Data Lab plugin for Claude Code",
  "version": "1.0.0"
}
EOF

# Run Claude Code with the plugin
claude --plugin-dir ./domino-claude-plugin

To make this persistent without the marketplace approach, add a shell alias:

echo 'alias claude="claude --plugin-dir /path/to/domino-claude-plugin"' >> ~/.bashrc
source ~/.bashrc

---

Option 3: Team / Project-Level Install

For teams sharing a project, add the marketplace to your project's .claude/settings.json:

{
  "extraKnownMarketplaces": {
    "domino-marketplace": {
      "source": {
        "source": "directory",
        "path": "/path/to/domino-marketplace"
      }
    }
  },
  "enabledPlugins": {
    "domino-claude-plugin@domino-marketplace": true
  }
}

When team members trust the repository folder, Claude Code will prompt them to install the marketplace and plugin automatically.

---

Verifying the Installation

After installation, test that the plugin is working:

1. Check slash commands are available:

   /domino-app-init

2. Test skill auto-invocation by asking a Domino-related question:

   Help me deploy a Streamlit app to Domino

Claude should automatically invoke the domino-app-deployment skill.

3. Check the plugin is listed:

   /plugin

Navigate to the Installed tab.

> Note: Plugin skills do not appear in the /skills list. They are auto-invoked by Claude based on task context and will show in Claude's init message at the top of a new session.

---

Updating the Plugin

If installed via the marketplace approach, navigate to the plugin source and pull updates:

cd ~/.claude/marketplaces/domino/plugins/domino-claude-plugin
git pull

Then restart Claude Code. If installed via --plugin-dir, pull updates in the cloned directory.

---

What's Included

Bundled MCP Server

The plugin includes a vendored copy of the Domino MCP Server that starts automatically when the plugin is enabled. It provides tools for running Domino jobs, checking job status/results, and syncing files with DFS-based projects.

• Inside a Domino workspace: Fully automatic — authentication uses ephemeral tokens, project info is auto-detected.
• Outside Domino (laptop): Set DOMINO_API_KEY and DOMINO_HOST as environment variables in your shell.

Requires uv to be installed (see Prerequisites).

Skills (20 Total)

| Skill | Description | | --- | --- | | domino-workspaces | Jupyter, VS Code, RStudio workspace management | | domino-jobs | Jobs and scheduled jobs execution | | domino-environments | Compute environments and Dockerfile customization | | domino-datasets | Data management, snapshots, and versioning | | domino-projects | Git integration and project collaboration | | domino-app-deployment | Deploy web apps (React, Streamlit, Dash) | | domino-experiment-tracking | MLflow experiment tracking and model registry | | domino-genai-tracing | @add_tracing decorator and DominoRun | | domino-model-endpoints | Deploy and call model APIs | | domino-model-monitoring | Drift detection and model quality tracking | | domino-flows | Flyte-based workflow orchestration | | domino-distributed-computing | Spark, Ray, Dask cluster management | | domino-ai-gateway | LLM proxy for OpenAI, Bedrock, etc. | | domino-launchers | Parameterized web forms for self-service | | domino-modeling-assistant | MCP server for AI-assisted model development | | domino-data-connectivity | S3 Mountpoint, AWS IRSA, Azure credentials | | domino-python-sdk | Python SDK (python-domino) and REST API | | domino-data-sdk | Data SDK (domino-data) for data sources, datasets, training sets | | domino-ui-design | Knowledge on Domino UI styling for integrated App design | | netapp-volumes | Enterprise-grade multi-terabyte NetApp ONTAP storage with near-instant snapshots and versioning |

Slash Commands

| Command | Description | | --- | --- | | /domino-app-init | Initialize a new Domino app with framework templates | | /domino-debug-proxy | Debug reverse proxy issues for apps | | /domino-experiment-setup | Set up MLflow experiment tracking | | /domino-trace-setup | Set up GenAI tracing with the Domino SDK |

Subagents

| Agent | Description | | --- | --- | | domino-deploy | Specialized agent for deploying apps, models, and endpoints | | domino-debug | Agent for debugging Domino issues and troubleshooting | | domino-setup | Agent for setting up new projects and configurations |

Output Styles

Switch output styles with /output-style:

| Style | Description | | --- | --- | | domino-learning | Educational mode with Domino Insights after each task | | domino-mlops | Production-focused with MLOps checklists and best practices |

---

Project Structure

domino-claude-plugin/
├── .claude-plugin/
│   └── plugin.json          # Plugin manifest
├── .mcp.json                # Bundled MCP server config (auto-starts)
├── mcp-servers/             # Vendored MCP servers
│   └── domino_mcp_server/   # Domino MCP Server (jobs, DFS sync)
├── agents/                  # Subagents
│   ├── domino-deploy.md
│   ├── domino-debug.md
│   └── domino-setup.md
├── output-styles/           # Custom output styles
│   ├── domino-learning.md
│   └── domino-mlops.md
├── skills/                  # 18 skill directories
│   ├── workspaces/
│   ├── jobs/
│   ├── environments/
│   ├── datasets/
│   ├── projects/
│   ├── app-deployment/
│   ├── experiment-tracking/
│   ├── genai-tracing/
│   ├── model-endpoints/
│   ├── model-monitoring/
│   ├── flows/
│   ├── distributed-computing/
│   ├── ai-gateway/
│   ├── launchers/
│   ├── modeling-assistant/
│   ├── data-connectivity/
│   ├── python-sdk/
│   ├── domino-data-sdk/
│   └── netapp-volumes/
├── commands/                # Slash commands
├── hooks/                   # Example automation hooks
├── templates/               # Code templates
│   ├── vite-react/
│   ├── streamlit/
│   ├── dash/
│   ├── experiment/
│   └── tracing/
├── CONTRIBUTING.md
├── LICENSE
└── README.md

---

Troubleshooting

| Issue | Solution | | --- | --- | | /skills shows "No skills found" | Plugin skills don't appear in /skills — they are auto-invoked by Claude based on context. Check /plugin → Installed tab instead. | | Plugin not loading from settings.json | Claude Code does not support a "plugins" array in settings.json. Use the marketplace approach or --plugin-dir flag. | | ~ path not expanding | Always use absolute paths (e.g., /home/ubuntu/...) in marketplace commands and settings. | | Slash commands not appearing | Restart Claude Code after installing. Commands are loaded at session start. | | "Failed to parse marketplace file" | Ensure marketplace.json has the owner object and source is a string path (e.g., "./plugins/domino-claude-plugin"), not a nested object. |

---

Usage Examples

Deploy a Streamlit App

User: Help me deploy a Streamlit dashboard to Domino
Claude: I'll help you set up a Streamlit app for Domino...

Set Up Experiment Tracking

User: /domino-experiment-setup
Claude: I'll configure MLflow experiment tracking for your project...

Create a Scheduled Job

User: How do I run a training script every day at midnight?
Claude: I'll show you how to create a scheduled job in Domino...

Deploy a Model API

User: I need to deploy my scikit-learn model as an API
Claude: I'll help you create a model endpoint in Domino...

---

API Reference

The domino-python-sdk skill includes comprehensive REST API documentation:

• API-PROJECTS.md — Projects, collaborators, Git repos
• API-JOBS.md — Jobs, logs, scheduled execution
• API-DATASETS.md — Datasets, snapshots, permissions
• API-MODELS.md — Model APIs, deployments, registry
• API-ENVIRONMENTS.md — Environments, revisions
• API-APPS.md — Apps, versions, instances
• API-ADMIN.md — Users, orgs, hardware tiers

---

Documentation

• Domino Documentation
• Domino API Guide
• Domino Blueprints
• python-domino GitHub
• Claude Code Plugin Docs
• Claude Code Marketplace Docs

Contributing

Contributions are welcome! Please feel free to submit issues and pull requests.

License

This project is licensed under the MIT License — see the LICENSE file for details.

Support

• For Domino platform issues: Domino Support
• For plugin issues: GitHub Issues