開発者向けドキュメント > DataRobot CLI > Template system > 環境変数

環境変数¶

This page outlines how to manage environment variables and .env files in DataRobot templates. DataRobot templates use .env files to store configuration variables needed by your application. The CLI provides tools to:

Create .env files from templates
Interactively edit variables
Validate configuration
Securely manage secrets

File structure¶

.env.template¶

The template provided by the repository (committed to Git):

# Required configuration
APP_NAME=
DATAROBOT_ENDPOINT=
DATAROBOT_API_TOKEN=

# Optional configuration
# DEBUG=false
# LOG_LEVEL=info
# PORT=8080

# Database configuration
# DATABASE_URL=
# DATABASE_POOL_SIZE=10

# Cache configuration
# CACHE_ENABLED=false
# CACHE_URL=

特性¶

Committed to version control
Contains empty required variables
Comments indicate optional variables
Includes documentation comments

.env¶

The actual configuration file (never committed):

# Required configuration
APP_NAME=my-awesome-app
DATAROBOT_ENDPOINT=https://app.datarobot.com
DATAROBOT_API_TOKEN=***

# Optional configuration
DEBUG=true
LOG_LEVEL=debug
PORT=8000

# Database configuration
DATABASE_URL=postgresql://localhost:5432/mydb
DATABASE_POOL_SIZE=5

Characteristics: - Generated from .env.template. - Contains actual values. - Never committed (in .gitignore). - User-specific configuration.

Create environment files¶

Use the wizard¶

The interactive wizard guides you through configuration.

# In a template directory
dr dotenv setup

または

# During template setup
dr templates setup

Wizard workflow¶

Loads .env.template.
Discovers configuration prompts.
Shows interactive questions.
Validates inputs.
Generates an .env file.

Manual creation¶

To copy and edit a template manually:

# Copy the template
cp .env.template .env

# Edit the template with your preferred editor
vim .env

# Alternatively, use the CLI editor
dr dotenv

Manage variables¶

Interactive editor¶

Launch the built-in editor to manage variables:

dr dotenv

特徴量¶

List all variables
Mask secrets (passwords, API keys)
Start wizard mode
Directly edit variables

Commands¶

Variables found in .env:

APP_NAME: my-awesome-app
DATAROBOT_ENDPOINT: https://app.datarobot.com
DATAROBOT_API_TOKEN: ***
DEBUG: true

Press w to set up variables interactively.
Press e to edit the file directly.
Press enter to finish and exit.

Wizard mode¶

You can also interactively configure a template with prompts.

dr dotenv setup

Advantages¶

Guided setup
Built-in validation
Conditional prompts
Help text for each variable

Direct editing¶

To edit the file directly:

dr dotenv edit
# Press 'e' to enter editor mode

# Or use external editor
vim .env

Variable types¶

Required variables¶

The following variables must be set before running the application:

# .env.template shows these without comments
APP_NAME=
DATAROBOT_ENDPOINT=
DATAROBOT_API_TOKEN=

The wizard enforces that an application name must be provided.

Enter your application name
> _
(Cannot proceed without entering a value)

Optional variables¶

The following variables are optional and can be left empty (shown as comments):

# .env.template shows these with # prefix
# DEBUG=false
# LOG_LEVEL=info

The wizard allows you to skip binding these variables:

Enable debug mode? (optional)
  > None (leave blank)
    Yes
    No

Secret variables¶

Sensitive values that should be masked during input and display.

To define secret variables:

# In .datarobot/prompts.yaml
prompts:
  - key: "api_key"
    env: "API_KEY"
    type: "secret_string"
    help: "Enter your API key"

Auto-detection¶

Variables with names containing PASSWORD, SECRET, KEY, or TOKEN are automatically treated as secrets.

Display behavior¶

The wizard input's secrets are masked with bullet characters (••••).
The editor view displays secrets as ***.
The actual file contains secrets as plain text values.

セキュリティのベストプラクティス¶

Always add .env to .gitignore.
Use secret_string type for all sensitive values.
Never commit .env files to version control.

Auto-generated secrets¶

You can cryptographically secure random values for application secrets:

prompts:
  - key: "session_secret"
    env: "SESSION_SECRET"
    type: "secret_string"
    generate: true
    help: "Session encryption key (auto-generated)"

特徴量¶

Generates 32-character random string if no value exists.
Uses base64 URL-safe encoding.
Preserves existing values (only generates when empty).
User can override secrets with a custom value.

Conditional variables¶

These variables are only shown or required based on your other selections:

# In .datarobot/prompts.yaml
prompts:
  - key: "enable_database"
    options:
      - name: "Yes"
        requires: "database_config"
      - name: "No"

  - key: "database_url"
    section: "database_config"
    env: "DATABASE_URL"
    help: "Database connection string"

If Enable database = No, then DATABASE_URL is not shown.

Environment variable discovery¶

The CLI discovers variables from multiple sources:

1. Template file (.env.template)¶

# Variables defined in template
APP_NAME=
PORT=8080

2. Prompt definitions (.datarobot/prompts.yaml)¶

prompts:
  - key: "app_name"
    env: "APP_NAME"
    help: "Application name"

3. Existing .env file¶

# Previously configured values
APP_NAME=my-app

4. Current environment¶

# Shell environment variables
export PORT=3000

Merge priority¶

The CLI merges in the following order of priority (highest priority first):

User input from wizard.
Current shell environment.
Existing .env values.
Template defaults.

Common patterns¶

Database configuration¶

# PostgreSQL
DATABASE_URL=postgresql://user:password@localhost:5432/dbname
DATABASE_POOL_SIZE=10
DATABASE_TIMEOUT=30

# MySQL
DATABASE_URL=mysql://user:password@localhost:3306/dbname

# MongoDB
DATABASE_URL=mongodb://localhost:27017/dbname

認証¶

# API Key
DATAROBOT_API_TOKEN=your_api_token_here

# OAuth
AUTH_PROVIDER=oauth2
AUTH_CLIENT_ID=client_id
AUTH_CLIENT_SECRET=***
AUTH_REDIRECT_URL=http://localhost:8080/callback

# JWT
JWT_SECRET=***
JWT_EXPIRATION=3600

Feature flags¶

# Enable/disable features
FEATURE_ANALYTICS=true
FEATURE_MONITORING=false
FEATURE_CACHING=true

# Or as comma-separated list
ENABLED_FEATURES=analytics,caching

ロギング¶

# Log level
LOG_LEVEL=debug  # debug, info, warn, error

# Log format
LOG_FORMAT=json  # json, text

# Log output
LOG_OUTPUT=stdout  # stdout, file

# Log file path
LOG_FILE=/var/log/app.log

セキュリティのベストプラクティス¶

Never commit .env files¶

Ensure that .gitignore includes:

# Environment variables
.env
.env.local
.env.*.local

# Keep templates
!.env.template
!.env.example

Use strong secrets¶

# ✓ Good - strong random secret
JWT_SECRET=a1b2c3d4e5f6g7h8i9j0k1l2m3n4o5p6

# ✗ Bad - weak secret
JWT_SECRET=secret123

Generate secure secrets:

# Random 32-byte hex string
openssl rand -hex 32

Restrict file permissions¶

# Only the owner can read/write
chmod 600 .env

# Verify
ls -la .env
# Should show: -rw------- (600)

Use different configs per environment¶

# Development
.env.development

# Staging
.env.staging

# Production
.env.production

Load based on the environment:

export ENV=production
dr run deploy

Avoid hardcoding in code¶

# ✗ Bad
api_token = "abc123"

# ✓ Good
import os
api_token = os.getenv("DATAROBOT_API_TOKEN")

検証¶

Validate with dr dotenv¶

To validate your environment configuration against template requirements:

dr dotenv validate

This command validates the following:

All required variables defined in .datarobot/prompts.yaml.
Core DataRobot variables (DATAROBOT_ENDPOINT, DATAROBOT_API_TOKEN).
Conditional requirements based on selected options.
Both .env file and environment variables.

Example output¶

Successful validation:

Validating required variables:
  APP_NAME: my-app
  DATAROBOT_ENDPOINT: https://app.datarobot.com
  DATAROBOT_API_TOKEN: ***
  DATABASE_URL: postgresql://localhost:5432/db

Validation passed: all required variables are set.

Validation errors:

Validating required variables:
  APP_NAME: my-app
  DATAROBOT_ENDPOINT: https://app.datarobot.com

Validation errors:

Error: required variable DATAROBOT_API_TOKEN is not set
  Description: DataRobot API token for authentication
  Set this variable in your .env file or run `dr dotenv setup` to configure it.

Error: required variable DATABASE_URL is not set
  Description: PostgreSQL database connection string
  Set this variable in your .env file or run `dr dotenv setup` to configure it.

ユースケース¶

Pre-flight checks before running tasks.
CI/CD pipeline validation.
Debugging missing configuration.
Troubleshooting application startup issues.

Required variables check¶

Commands like dr run automatically validate required variables.

$ dr run dev
Error: Missing required environment variables:
  - APP_NAME
  - DATAROBOT_API_TOKEN

Please run: dr dotenv setup

Format validation¶

For variables with specific formats:

# URL validation
DATAROBOT_ENDPOINT=https://app.datarobot.com  # ✓ Valid
DATAROBOT_ENDPOINT=not-a-url                   # ✗ Invalid

# Port validation
PORT=8080    # ✓ Valid
PORT=99999   # ✗ Invalid (out of range)

# Email validation
EMAIL=user@example.com  # ✓ Valid
EMAIL=invalid           # ✗ Invalid

Advanced features¶

Variable substitution¶

Reference other variables:

# Base URL
BASE_URL=https://app.datarobot.com

# API endpoint uses base URL
API_ENDPOINT=${BASE_URL}/api/v2

# Full URL becomes: https://app.datarobot.com/api/v2

Multi-line values¶

For long values:

# Single line
PRIVATE_KEY="-----BEGIN PRIVATE KEY-----\nMIIE..."

# Or use actual newlines
PRIVATE_KEY="-----BEGIN PRIVATE KEY-----
MIIEvQIBADANBgkqhkiG9w0BAQEFAASCBKcwggSjAgEAAoIBAQC...
-----END PRIVATE KEY-----"

コメント¶

Document your configuration:

# Application Configuration
APP_NAME=my-app          # Application identifier
PORT=8080                # HTTP server port

# Database Configuration
# Format: protocol://user:password@host:port/database
DATABASE_URL=postgresql://localhost:5432/mydb

トラブルシューティング¶

Variables not loading¶

# Check .env exists
ls -la .env

# Verify format
cat .env

# Check for syntax errors
# Each line should be: KEY=value

Secrets exposed¶

# Check .gitignore includes .env
cat .gitignore | grep .env

# Check Git status
git status
# Should NOT show .env

# If .env is tracked, remove it
git rm --cached .env
git commit -m "Remove .env from tracking"

Permission errors¶

# Fix permissions
chmod 600 .env

# Verify
ls -la .env

Variables not expanding¶

# Ensure proper syntax for variable substitution
# Works:
API_URL=${BASE_URL}/api

# Doesn't work:
API_URL=$BASE_URL/api  # Missing braces

Configuration not working¶

Use dr dotenv validate to diagnose issues:

# Validate configuration
dr dotenv validate

# If validation passes but issues persist, check:
# 1. Environment variables override .env
env | grep DATAROBOT

# 2. Ensure .env is in correct location (repository root)
pwd
ls -la .env

# 3. Check if application is loading .env file
# Some applications need explicit .env loading

一般的なワークフロー¶

Initial setup¶

cd my-template
dr dotenv setup
dr dotenv validate
dr run dev

Update credentials¶

dr auth login
dr dotenv update
dr dotenv validate

Validate before deployment¶

dr dotenv validate && dr run deploy

Edit and validate¶

dr dotenv edit
dr dotenv validate

環境変数¶

File structure¶

.env.template¶

特性¶

.env¶

Create environment files¶

Use the wizard¶

Wizard workflow¶

Manual creation¶

Manage variables¶

Interactive editor¶

特徴量¶

Commands¶

Wizard mode¶

Advantages¶

Direct editing¶

Variable types¶

Required variables¶

Optional variables¶

Secret variables¶

Auto-detection¶

Display behavior¶

セキュリティのベストプラクティス¶

Auto-generated secrets¶

特徴量¶

Conditional variables¶

Environment variable discovery¶

1. Template file (.env.template)¶

2. Prompt definitions (.datarobot/prompts.yaml)¶

3. Existing .env file¶

4. Current environment¶

Merge priority¶

Common patterns¶

Database configuration¶

認証¶

Feature flags¶

ロギング¶

セキュリティのベストプラクティス¶

Never commit .env files¶

Use strong secrets¶

Restrict file permissions¶

Use different configs per environment¶

Avoid hardcoding in code¶

検証¶

Validate with dr dotenv¶

Example output¶

ユースケース¶

Required variables check¶

Format validation¶

Advanced features¶

Variable substitution¶

Multi-line values¶

コメント¶

トラブルシューティング¶

Variables not loading¶

Secrets exposed¶

Permission errors¶

Variables not expanding¶

Configuration not working¶

一般的なワークフロー¶

Initial setup¶

Update credentials¶

Validate before deployment¶

Edit and validate¶

See also¶