The Omium CLI (omium) provides commands for configuration, running workflows, and managing executions.
Quick Reference
| Command | Description |
|---|
omium init | Initialize SDK configuration |
omium run <script.py> | Run script with tracing |
omium list | List recent executions |
omium show <id> | Show execution details |
omium logs <id> | Stream execution logs |
omium replay <id> | Replay from checkpoint |
omium tui | Launch Terminal UI |
Installation
The CLI is automatically installed with the SDK.
pip install omium
omium --version # Verify installation
Configuration Commands
omium init
Initialize Omium SDK configuration. This is the first command to run.
Options:
--api-key TEXT - API key (recommended for Windows/PowerShell)--api-url TEXT - API URL (default: https://api.omium.ai)--skip-verify - Skip API key verification
Example:
# Interactive mode
omium init
# Non-interactive (Windows recommended)
omium init --api-key om_xxxxx
omium configure [OPTIONS]
--api-key TEXT - New API key--api-url TEXT - API URL--region TEXT - Region (us-east-1, eu-west-1)--interactive / --no-interactive - Interactive prompts
Workflow Commands
omium run
Run a Python script or workflow with Omium instrumentation.
omium run <script.py|workflow.json> [OPTIONS]
--project TEXT - Project name--env KEY=VALUE - Environment variables (can repeat)--no-trace - Disable automatic tracing--execution-id TEXT - Custom execution ID
Examples:
# Run Python script with auto-instrumentation
omium run my_agent.py
# Run with environment variables
omium run my_agent.py --env OPENAI_API_KEY=sk-xxx --env DEBUG=true
# Run workflow JSON (legacy)
omium run workflow.json
omium init-workflow
Create a workflow template file.
omium init-workflow --type <TYPE> --name <NAME>
--type TEXT - Workflow type: crewai, langgraph, basic--name TEXT - Workflow name--output TEXT - Output file path
Example:
omium init-workflow --type langgraph --name my-agent
Execution Commands
omium list
List recent executions.
Options:
--status TEXT - Filter by status (running, completed, failed)--workflow-id TEXT - Filter by workflow--limit INT - Number of results (default: 20)
Example:
omium list
omium list --status failed --limit 10
omium show
Show detailed information about an execution.
omium show <EXECUTION_ID>
omium logs
Stream logs from a running execution.
omium logs <EXECUTION_ID> [OPTIONS]
--follow / --no-follow - Stream live logs (default: follow)-n, --tail INT - Number of recent logs to show
Examples:
omium logs exec-123 # Stream live
omium logs exec-123 --no-follow # Show last logs and exit
omium logs exec-123 -n 100 # Show last 100 logs
omium replay
Replay an execution from a checkpoint.
omium replay <EXECUTION_ID> [OPTIONS]
--checkpoint-id TEXT - Specific checkpoint to replay from
Example:
omium replay exec-abc123
omium replay exec-abc123 --checkpoint-id cp-456
Project Commands
omium push
Push local workflow/project to Omium cloud.
Options:
--project TEXT - Project directory--name TEXT - Workflow name
TUI Mode
omium tui
Launch the interactive Terminal UI for browsing executions.
Environment Variables
| Variable | Description |
|---|
OMIUM_API_KEY | API key (overrides config file) |
OMIUM_API_URL | API URL |
OMIUM_DEBUG | Enable debug logging |
Configuration File
Config is stored at ~/.omium/config.json:
{
"api_key": "om_xxx",
"api_url": "https://api.omium.ai/api/v1",
"region": "us-east-1",
"default_project": null
}
Next Steps
