Command Line Interface (CLI)

Use Klyra from your terminal

Overview

The Klyra CLI provides a convenient way to interact with the Klyra API directly from your terminal. It’s useful for testing, scripting, batch processing, and CI/CD pipelines.

Installation

npm

npm install -g @klyra/cli

Verify installation:

klyra --version

Authentication

Authenticate with your API key using one of these methods:

# Login interactively
klyra login

# Set API key directly
klyra config set api_key YOUR_API_KEY

# Use environment variable
export KLYRA_API_KEY=YOUR_API_KEY

Basic Commands

Moderate Text Content

# Moderate text from a string
klyra moderate text "This is a text to moderate"

# Moderate text from a file
klyra moderate text --file ./content.txt

# Specify categories to check
klyra moderate text --file ./content.txt --categories toxic,harassment,spam

# Set custom threshold
klyra moderate text --file ./content.txt --threshold 0.8

Moderate Image Content

# Moderate an image from a file
klyra moderate image --file ./image.jpg

# Moderate an image from a URL
klyra moderate image --url https://example.com/image.jpg

# Get detailed analysis with region detection
klyra moderate image --file ./image.jpg --detailed

Moderate Audio/Video Content

# Moderate audio
klyra moderate audio --file ./audio.mp3

# Moderate video
klyra moderate video --file ./video.mp4 --frame-rate 1

Batch Processing

# Moderate multiple files
klyra moderate batch --glob "./content/*.txt" --output results.json

# Process a directory of images
klyra moderate batch --dir ./images --type image --output results.csv

# Process with custom settings
klyra moderate batch --glob "./content/*.txt" --settings settings.json

Output Formats

Control the output format:

# JSON output (default)
klyra moderate text "Hello world" --format json

# Compact JSON (single line)
klyra moderate text "Hello world" --format json-compact

# CSV output
klyra moderate text "Hello world" --format csv

# Pretty table output
klyra moderate text "Hello world" --format table

# Colored output for human-readable results
klyra moderate text "Hello world" --format pretty

Configuration

View and edit your CLI configuration:

# View current configuration
klyra config list

# Set a configuration value
klyra config set <key> <value>

# Get a configuration value
klyra config get <key>

# Reset to default configuration
klyra config reset

Available configuration options:

Option	Description	Default
api_key	Your Klyra API key	-
api_url	API base URL	https://api.klyra.io/v1
output_format	Default output format	json
timeout	Request timeout in seconds	60
max_retries	Maximum retry attempts	3
color	Enable/disable colored output	true

Advanced Usage

Using Settings Files

Create a JSON settings file for reusable configurations:

{
  "categories": ["toxic", "harassment", "spam"],
  "threshold": 0.7,
  "language": "en",
  "detailed": true
}

Use the settings file:

klyra moderate text --file ./content.txt --settings ./settings.json

CI/CD Integration

Use the CLI in your CI/CD pipelines:

# Example GitHub Actions workflow
klyra moderate batch --glob "./content/**/*.txt" --format json --output results.json --exit-code

# Exit with error if any content is flagged
if [ $? -eq 1 ]; then
  echo "Moderation failed - inappropriate content detected"
  exit 1
fi

Webhooks

Manage webhooks from the CLI:

# List all webhooks
klyra webhooks list

# Create a new webhook
klyra webhooks create --url https://example.com/webhook --events moderation.completed,moderation.flagged

# Delete a webhook
klyra webhooks delete <webhook_id>

Custom Headers

Add custom headers to requests:

klyra moderate text "Hello" --header "Custom-Header: value" --header "Another-Header: value"

Plugins

Extend the CLI with plugins:

# Install a plugin
klyra plugins install @klyra/cli-plugin-jira

# Use plugin functionality
klyra jira create-issue --moderation-id mod_123456

# List installed plugins
klyra plugins list

# Remove a plugin
klyra plugins uninstall @klyra/cli-plugin-jira

Trustless Mode

Use end-to-end encryption with trustless mode:

# Generate an encryption key
klyra trustless generate-key --output key.txt

# Use trustless mode with a specified key
klyra moderate text "Sensitive content" --trustless --key-file ./key.txt

# Use trustless mode with an environment variable
export KLYRA_TRUSTLESS_KEY=$(cat key.txt)
klyra moderate text "Sensitive content" --trustless

Examples

Content Moderation Pipeline

#!/bin/bash
# Process user uploads and flag inappropriate content

# Set up error handling
set -e

# Process all new uploads
klyra moderate batch \
  --glob "./uploads/*.{jpg,png,gif}" \
  --type image \
  --categories adult,violence,hate_symbols \
  --threshold 0.7 \
  --output results.json \
  --format json

# Extract flagged content
cat results.json | jq '.[] | select(.flagged == true) | .file' > flagged_files.txt

# Take action on flagged content
if [ -s flagged_files.txt ]; then
  echo "Inappropriate content detected, sending notification..."
  # Send notification or trigger moderation workflow
fi

Interactive Mode

The CLI also offers an interactive mode:

klyra interactive

This opens an interactive session where you can enter commands and see results immediately, perfect for exploration and testing.

Support

For more information and support:

• Run klyra --help or klyra <command> --help for command-specific help
• Visit the CLI documentation for detailed guides
• Report issues on GitHub
• Contact [e-mail] for assistance