xase-ai 2.0.0

Enterprise-grade CLI for the Xase AI Lab (governed AI training datasets)

🎨 Xase AI Lab CLI

Enterprise-grade command-line interface for governed AI training datasets

Beautiful, interactive CLI for discovering, managing, and consuming governed AI training datasets with enterprise compliance built-in.

✨ Features

• 🎨 Beautiful UI - Claude-inspired orange theme with rich terminal formatting
• 🚀 Interactive Onboarding - Guided setup wizard for first-time users
• 🔐 Secure Authentication - Email OTP with token management
• 📊 Rich Tables - Formatted displays for offers, leases, and usage stats
• ⚡ Fast & Reliable - Built with async support and retry logic
• 🛡️ Compliance-First - GDPR & AI Act compliant data access

📦 Installation

# Clone or navigate to the CLI directory
cd packages/xase-cli

# Create virtual environment
python3 -m venv .venv
source .venv/bin/activate

# Install in editable mode
pip install -e .

# Verify installation
xase-cli --help

🚀 Quick Start

First Run - Interactive Onboarding

On first run, Xase CLI will guide you through setup:

xase-cli

You'll see:

1. Beautiful ASCII art banner with orange theme
2. Interactive setup wizard to configure API URL
3. Guided next steps for authentication and usage

Manual Setup

Run the setup wizard anytime:

xase-cli setup

Choose from:

• 🌐 API Connection Settings
• 🔐 Authentication & Login
• ⚙️ Advanced Settings
• 👁️ View Current Configuration

Authentication

# Login with email OTP
xase-cli login
# Enter your email
# Check email for 6-digit code
# Enter code to complete authentication

# View usage statistics
xase-cli usage

# Logout
xase-cli logout

Working with Datasets

# List available offers
xase-cli list-offers --limit 20 --risk LOW

# Create a lease for a dataset
xase-cli mint-lease DATASET_ID --ttl-seconds 3600

# View active leases
xase-cli list-leases

# Get lease details
xase-cli lease-details LEASE_ID

# Stream dataset for training
xase-cli stream DATASET_ID --lease-id LEASE_ID --output batch.json

📚 Commands Reference

Core Commands

setup

Interactive setup wizard

xase-cli setup

login

Authenticate via email OTP

xase-cli login

logout

Remove saved tokens

xase-cli logout

version

Show CLI version

xase-cli version

Dataset Discovery

list-offers

List available access offers with filters

xase-cli list-offers [OPTIONS]

Options:
  --limit N          Maximum number of offers (default: 20)
  --risk LEVEL       Filter by risk level (LOW, MEDIUM, HIGH)
  --language LANG    Filter by language

Example:

xase-cli list-offers --risk MEDIUM --language en --limit 50

---

Lease Management

mint-lease

Create a new lease for a dataset

xase-cli mint-lease DATASET_ID [OPTIONS]

Arguments:
  DATASET_ID         Dataset ID to lease

Options:
  --ttl-seconds N    Lease time-to-live in seconds (default: 1800)

Example:

xase-cli mint-lease ds_abc123 --ttl-seconds 3600

Output:

✓ Lease minted
ℹ Lease ID: lease_xyz789
ℹ Expires: 2026-02-15T19:30:00Z

---

list-leases

List active leases

xase-cli list-leases [OPTIONS]

Options:
  --limit N          Maximum number of leases (default: 20)

Example:

xase-cli list-leases --limit 10

Output:

                Active Leases                
╭──────────────┬────────────┬────────┬─────────╮
│ Lease ID     │ Dataset    │ Status │ Expires │
├──────────────┼────────────┼────────┼─────────┤
│ lease_xyz123 │ ds_med_001 │ ACTIVE │ 18:30   │
╰──────────────┴────────────┴────────┴─────────╯

---

lease-details

Show detailed information about a lease

xase-cli lease-details LEASE_ID

Arguments:
  LEASE_ID           Lease ID to inspect

Example:

xase-cli lease-details lease_xyz789

Data Streaming

stream

Stream dataset batch for training

xase-cli stream DATASET_ID [OPTIONS]

Arguments:
  DATASET_ID         Dataset to stream

Options:
  --lease-id ID      Lease ID to use (required)
  --env ENV          Environment (development, staging, production)
  --estimated-hours  Estimated hours of usage (default: 0.5)
  --output FILE      Output file path

Example:

xase-cli stream ds_def456 \
  --lease-id lease_xyz789 \
  --env production \
  --output training_batch.json

---

Monitoring

usage

Show usage statistics

xase-cli usage

Output:

       Usage Statistics        
╭───────────────┬───────────╮
│ Tenant ID     │ tenant_01 │
│ Active Offers │        15 │
│ Active Leases │         3 │
│ Auth Mode     │   api_key │
╰───────────────┴───────────╯

---

🔄 Production Workflow

Complete flow from discovery to consumption:

# 1. First-time setup
xase-cli setup
xase-cli login

# 2. Discover datasets
xase-cli list-offers --risk MEDIUM --language en

# 3. Create lease for dataset
xase-cli mint-lease ds_abc123 --ttl-seconds 7200
# Note the lease ID from output

# 4. Stream data in batches
xase-cli stream ds_abc123 \
  --lease-id lease_xyz789 \
  --env production \
  --output batch_001.json

xase-cli stream ds_abc123 \
  --lease-id lease_xyz789 \
  --env production \
  --output batch_002.json

# 5. Monitor active leases
xase-cli list-leases

# 6. Check usage statistics
xase-cli usage

# 7. View lease details
xase-cli lease-details lease_xyz789

---

Error Handling

The CLI exits with standard codes:

• 0 - Success
• 1 - General error (API error, validation failed)
• 2 - Configuration error (missing API key)
• 130 - Interrupted by user (Ctrl+C)

Example error:

$ python xase_cli.py validate pol_invalid
✗ HTTP 404: Policy not found

---

Best Practices

1. Always validate before consuming

# ✅ Good
python xase_cli.py validate pol_xyz789 --requested-hours 0.5
python xase_cli.py stream ds_def456 --output data.json

# ❌ Bad (no validation)
python xase_cli.py stream ds_def456 --output data.json

2. Use small batches for long-running jobs

# ✅ Good - consume in 0.5h increments
for i in {1..10}; do
  python xase_cli.py validate pol_xyz789 --requested-hours 0.5
  python xase_cli.py stream ds_def456 --output batch_$(printf %03d $i).json
  sleep 60
done

# ❌ Bad - single large request
python xase_cli.py stream ds_def456 --output all_data.json  # May timeout

3. Monitor lease expiration

# Check lease status regularly
python xase_cli.py list-leases | grep "Expires"

4. Secure your API key

# ✅ Good - use environment variable
export XASE_API_KEY="xase_pk_..."

# ❌ Bad - hardcode in script
XASE_API_KEY="xase_pk_..." python xase_cli.py list-offers

---

Automation Example

Bash script for continuous consumption:

#!/bin/bash
set -e

POLICY_ID="pol_xyz789"
DATASET_ID="ds_def456"
BATCH_SIZE=0.5
OUTPUT_DIR="./training_data"

mkdir -p "$OUTPUT_DIR"

for i in {1..20}; do
  echo "Processing batch $i..."
  
  # Validate access
  python xase_cli.py validate "$POLICY_ID" --requested-hours "$BATCH_SIZE"
  
  # Stream data
  OUTPUT_FILE="$OUTPUT_DIR/batch_$(printf %03d $i).json"
  python xase_cli.py stream "$DATASET_ID" --output "$OUTPUT_FILE"
  
  echo "✓ Batch $i saved to $OUTPUT_FILE"
  
  # Wait before next batch
  sleep 30
done

echo "✓ All batches completed"

---

Troubleshooting

"XASE_API_KEY environment variable not set"

Fix:

export XASE_API_KEY="xase_pk_..."

"HTTP 401: Unauthorized"

Causes:

• Invalid API key
• Expired API key
• Wrong tenant

Fix:

• Verify API key in dashboard
• Regenerate if expired

"HTTP 403: Access denied"

Causes:

• Policy expired
• Usage limit exceeded
• Lease revoked

Fix:

# Check policy status
python xase_cli.py validate pol_xyz789

# Request new lease if needed
python xase_cli.py execute off_abc123 --hours 1.0

"HTTP 429: Rate limit exceeded"

Fix:

• Add delays between requests
• Reduce batch frequency
• Contact support to increase rate limit

---

Support

• Documentation: docs.xase.ai
• API Reference: api.xase.ai/docs
• Issues: github.com/xase/xase-cli/issues
• Email: [e-mail]

---

License

MIT © Xase

---

Built for AI Labs by the Xase team