Dagster CLI (dgc)

A command-line interface for Dagster+, inspired by GitHub's gh CLI.

Installation

# Install with uv (recommended - adds 'dgc' to PATH)
uv tool install dagster-cli

# Or install with pip
pip install dagster-cli

Quick Start

# Authenticate with your Dagster+ deployment
dgc auth login

# Check health status of all assets (find problems)
dgc asset health

# View details about a specific asset
dgc asset view analytics/daily_revenue

# View logs for a failed run (see stack traces)
dgc run logs abc123 --stderr

# List recent runs to find failures
dgc run list --limit 10

Note: All commands support --help for detailed options and --tldr for quick examples.

MCP Server for AI Assistants

Run the MCP server without installation using uvx:

# Start MCP server for Claude Code integration
uvx --from dagster-cli dgc mcp start

Configure Claude Code by adding to your MCP settings:

{
  "mcpServers": {
    "dagster-cli": {
      "command": "uvx",
      "args": ["--from", "dagster-cli", "dgc", "mcp", "start"]
    }
  }
}

This enables AI assistants to:

• Check asset health and identify failures
• Investigate failed runs and view error logs
• Monitor job execution and debug issues
• Access stderr/stdout logs for troubleshooting

Working with Branch Deployments

# List all deployments including branches
dgc deployment list

# Get filtered logs from a branch deployment (e.g., for PR review)
dgc run list --deployment feat-new-feature --status FAILURE --limit 5
dgc run logs abc123 --deployment feat-new-feature --stderr

# Filter runs by job name on a specific deployment
dgc run list --deployment staging --job daily_etl --limit 10

Features

• Secure Authentication - Store credentials safely with profile support
• Job Management - List, view, and run Dagster jobs from the terminal
• Run Monitoring - Track run status, view logs, and analyze failures
• Asset Management - List, materialize, and monitor asset health
• Repository Operations - List and reload code locations
• Profile Support - Manage multiple Dagster+ deployments
• Branch Deployment Support - Access branch deployments for testing and debugging
• Deployment Discovery - List and test available deployments
• MCP Integration - AI assistant integration for monitoring and debugging

Configuration

Authentication

dgc auth login                  # Set up credentials
dgc auth status                 # View current profile
dgc auth switch staging         # Switch between profiles

Multiple Profiles

dgc auth login --profile staging    # Create new profile
dgc job list --profile production   # Use specific profile

Environment Variables

• DAGSTER_CLOUD_TOKEN - User token
• DAGSTER_CLOUD_URL - Deployment URL
• DGC_PROFILE - Default profile
• DAGSTER_CLOUD_DEPLOYMENT - Default deployment

Credentials stored in ~/.config/dagster-cli/config.json

Development

# Run tests
uv run pytest

# Format and lint
make fix

# Build package
uv build