Interactive Exploration

Learn how to use Ona Terminal effectively through hands-on interaction.

Using / Commands (Interactive Mode)

Master Ona Terminal’s interactive CLI with slash commands and examples.

Getting Started

Launch interactive mode:

ona-terminal

You’ll see the interactive prompt:

🤖 |

Essential Commands

System Commands:

# Show all available commands
🤖 | /help

# List all slash commands
🤖 | /commands

# Check system status and model availability
🤖 | /status

# Show available AI models
🤖 | /models

# Exit the CLI
🤖 | /exit

Code Generation:

# Generate code with natural language
🤖 | generate a python function that reads CSV files
🤖 | create a terraform module for AWS S3 bucket
🤖 | write unit tests for my authentication function

# Generate with specific complexity
🤖 | /generate --complexity high --language rust "HTTP client with retry logic"

Custom Model Commands:

# Use specific fine-tuned model
🤖 | /model mistral-policy-analysis "analyze economic impact of solar policy"

# Use IaC specialized model
🤖 | /model mistral-iac-generation "create terraform for multi-AZ deployment"

# Use Qwen Claude-MD model for complex reasoning
🤖 | /model qwen-claude-md "complex architectural design with CLAUDE.md methodology"

Understanding Agents

Learn how agents work in Ona Terminal and how to create custom single-use agents.

What Are Agents?

Agents in Ona Terminal are specialized AI components that perform specific tasks. Each agent follows the single responsibility principle - doing one thing well.

Types of Agents

1. Built-in Agents:

# Code generation agent
🤖 | /generate "python function to parse JSON"

# Documentation agent
🤖 | /docs "explain AWS Lambda best practices"

# Analysis agent
🤖 | /analyze "review this terraform module for security issues"

2. Model-Specific Agents:

# Use a specific model as an agent
🤖 | /model claude-3-sonnet "complex reasoning task"
🤖 | /model mistral-7b "generate infrastructure code"

3. Custom Single-Use Agents:

# Define your own agent for specific workflows
🤖 | /agent create data-validator --task "validate CSV format"
🤖 | /agent create cost-analyzer --task "analyze AWS spending"

Creating Custom Agents

Agent Definition:

# custom-agents.yaml
agents:
  log-analyzer:
    description: "Analyzes application logs for errors"
    model: "claude-3-haiku"
    system_prompt: "You are a log analysis expert..."
    
  api-tester:
    description: "Tests API endpoints and reports issues"
    model: "mistral-7b"
    system_prompt: "You test APIs systematically..."

Using Custom Agents:

# Invoke custom agent
🤖 | /log-analyzer --file app.log --severity error

# Chain agents together
🤖 | /api-tester --endpoint /users | /log-analyzer

Agent Orchestration

Agents can work together in workflows:

# Sequential execution
🤖 | /analyze code.py > /generate tests > /validate

# Parallel execution
🤖 | /parallel --agents "analyzer,linter,security-scan" --target src/

# Conditional execution
🤖 | /if-error /analyze > /debug > /fix

Loading Bedrock & Custom Models

Configure AWS Bedrock and deploy custom fine-tuned models.

AWS Bedrock Configuration

Recommended Models:

Amazon Nova Pro - Primary model for best availability
Amazon Nova Lite - Fast fallback option

Setup Process: Configure AWS credentials and region for Bedrock access. Ona Terminal will automatically detect available models in your region.

Custom Fine-Tuned Models

Available Models (from deployments):

🎯 Mistral Policy Analysis (7B):

# Deploy model
🤖 | /deploy-model --name mistral-policy-analysis --instance g5.2xlarge

# Use for analysis
🤖 | /model mistral-policy-analysis "economic impact analysis of renewable energy policy"

🏗️ Mistral Infrastructure-as-Code (7B):

# Deploy model
🤖 | /deploy-model --name mistral-iac-generation --instance g5.2xlarge

# Generate infrastructure
🤖 | /model mistral-iac-generation "terraform module for secure multi-region deployment"

🧠 Qwen Claude-MD (14B):

# Deploy larger model
🤖 | /deploy-model --name qwen-claude-md --instance g5.4xlarge

# Complex reasoning tasks
🤖 | /model qwen-claude-md "architectural design following CLAUDE.md explore-plan-code-commit methodology"

Model Deployment Commands

Check Model Availability:

# List deployed models
🤖 | /models --deployed

# Check model health
🤖 | /model-health --all

# Show deployment status
🤖 | /deployment-status

Deploy Custom Models:

# Deploy single model
🤖 | /deploy --model mistral-policy-analysis --instance g5.2xlarge

# Deploy unified Mistral (60% cost savings)
🤖 | /deploy --unified mistral-dual --instance g5.2xlarge

# Scale model deployment
🤖 | /scale --model qwen-claude-md --instances 2

Cost Optimization Strategies

Unified Deployment (60% savings):

# Deploy both Mistral models on single instance
🤖 | /deploy --unified mistral-dual
# Result: IaC + Policy models sharing base Mistral-7B

Instance Sizing:

# Cost-effective for Mistral models
🤖 | /deploy --model mistral-iac-generation --instance g4dn.xlarge

# Performance-optimized for Qwen
🤖 | /deploy --model qwen-claude-md --instance g5.4xlarge

What’s Next?

See Real-World Implementation - Complete O&M business case
Technical Setup - 5-minute developer setup
Business Value - Understanding ROI and benefits

Explore O&M Use Case

Get Help & Stay Updated

Contact Support

For technical assistance, feature requests, or any other questions, please reach out to our dedicated support team.

Email Support Join Our Discord