orq-cli 0.1.4

Unofficial CLI tool for orq.ai LLM Ops platform

ORQ CLI

Command line interface for the orq.ai LLM Ops platform.

Installation

pip install orq-cli

Quick Start

Configure API Key

Set your API key using one of these methods:

# Using the CLI
orq config set api_key YOUR_API_KEY

# Using environment variable
export ORQ_API_KEY=YOUR_API_KEY

Basic Usage

# List deployments
orq deployments list

# Invoke a deployment with a message
orq invoke my-deployment -m "Hello, world!"

# Invoke with variables (for prompt templates)
orq invoke my-deployment -v firstname=John -v city=Paris

# Combine variables and message
orq invoke my-deployment -v customer=John -m "Help me with my order"

# Stream a response
orq invoke my-deployment -v topic=AI --stream

# Interactive mode
orq --interactive

Passing Variables

When your deployment has prompt templates with placeholders like {{firstname}} or {{city}}, pass them using -v or --var:

# Using --var (recommended)
orq invoke customer-service -v firstname=John -v city=Paris -v language=English

# Multiple variables
orq deployments invoke my-key -v name=Alice -v topic="machine learning" -v format=markdown

# Using JSON (alternative)
orq deployments invoke my-key --inputs '{"firstname": "John", "city": "Paris"}'

Commands

Deployments

orq deployments list                              # List all deployments
orq deployments invoke KEY -v name=value          # Invoke with variables
orq deployments stream KEY -v name=value          # Stream with variables
orq deployments get-config KEY                    # Get deployment configuration

Datasets

orq datasets list                       # List datasets
orq datasets create --name "My Dataset" # Create a dataset
orq datasets get DATASET_ID             # Get dataset details
orq datasets delete DATASET_ID          # Delete a dataset

# Datapoints
orq datasets datapoints list DATASET_ID
orq datasets datapoints create DATASET_ID --inputs '{"key": "value"}'

Files

orq files list                          # List files
orq files upload path/to/file.txt       # Upload a file
orq files get FILE_ID                   # Get file details
orq files delete FILE_ID                # Delete a file

Knowledge Bases

orq knowledge list                      # List knowledge bases
orq knowledge search KB_ID --query "search term"
orq knowledge datasources list KB_ID    # List datasources
orq knowledge chunks list KB_ID DS_ID   # List chunks

Prompts

orq prompts list                        # List prompts
orq prompts get PROMPT_ID               # Get prompt details

Contacts

orq contacts create EXTERNAL_ID --name "John Doe" --email "[e-mail]"

Feedback

orq feedback create TRACE_ID --field rating --value 5

Configuration

orq config show                         # Show configuration
orq config set KEY VALUE                # Set a config value
orq config get KEY                      # Get a config value
orq config path                         # Show config file path

Quick Shortcuts

# Quick list any resource
orq list deployments
orq list datasets
orq list files

# Quick invoke with message
orq invoke DEPLOYMENT_KEY -m "message"

# Quick invoke with variables
orq invoke DEPLOYMENT_KEY -v firstname=John -v city=Paris

# Quick invoke with streaming
orq invoke DEPLOYMENT_KEY -v topic=AI -m "Explain this" --stream

Interactive Mode

Run the CLI in interactive mode for a guided experience:

orq --interactive
# or
orq -i

Output Formats

All commands support different output formats:

orq deployments list --output table   # Default, pretty table
orq deployments list --output json    # JSON output
orq deployments list --output yaml    # YAML output

Configuration File

Configuration is stored in ~/.orq/config.yaml:

api_key: your-api-key
environment: production
output_format: table

Environment Variables

• ORQ_API_KEY: API key for authentication
• ORQ_ENVIRONMENT: Default environment (production, staging, etc.)

Development

# Install in development mode
pip install -e ".[dev]"

# Run tests
pytest

# Run tests with coverage
pytest --cov=orq