Configuration Overview

The Rawi Command Line Interface (Rawi CLI) uses configuration settings to determine how it interacts with AI providers and manages your sessions. Proper configuration ensures optimal performance and seamless integration with your workflow.

Configuration Components

Rawi configuration includes:

Credentials: API keys and authentication tokens for AI providers
Provider and Model: Which AI service and specific model to use
Profiles: Named configurations for different use cases
Generation Settings: Temperature, max tokens, and language preferences
Provider Settings: Custom endpoints, regions, and provider-specific options

Configuration Precedence

Rawi CLI reads configuration from multiple sources in order of precedence:

Command line options – Highest priority (e.g., --provider, --model, --api-key)
Environment variables – System-level settings (e.g., OPENAI_API_KEY, RAWI_PROFILE)
Profile configuration files – User-defined profiles in ~/.config/rawi/profiles/
Default profile – Fallback configuration used when no profile is specified

Run interactive configuration:
Terminal window
```
rawi configure
```
This will guide you through:
- Selecting an AI provider
- Choosing a model
- Entering credentials
- Setting preferences
Verify your setup:
Terminal window
```
rawi info
```
Test your configuration:
Terminal window
```
rawi ask "Hello, world!"
```

Configure OpenAI provider:

rawi configure --provider openai --model gpt-4o --api-key sk-your-key

Set generation parameters:

rawi configure --temperature 0.7 --max-tokens 2048 --language english

Test the configuration:

rawi ask "What is artificial intelligence?"

Profile Management

Profiles allow you to maintain different configurations for various use cases:

Creating Profiles

# Work profile with enterprise settings
rawi configure --profile work \
  --provider azure \
  --model gpt-4 \
  --resource-name company-openai \
  --api-key work-key

# Personal profile for experimentation
rawi configure --profile personal \
  --provider openai \
  --model gpt-3.5-turbo \
  --api-key personal-key \
  --temperature 0.8

# Local profile for privacy
rawi configure --profile local \
  --provider ollama \
  --model llama2 \
  --base-url http://localhost:11434

Profile Examples by Use Case

Development Profiles

# Code review profile (detailed analysis)
rawi configure --profile code-review \
  --provider anthropic \
  --model claude-3-5-sonnet-20241022 \
  --temperature 0.1 \
  --max-tokens 4000

# Quick development profile (fast responses)
rawi configure --profile dev-quick \
  --provider openai \
  --model gpt-3.5-turbo \
  --temperature 0.3 \
  --max-tokens 1000

Content Creation Profiles

# Creative writing profile
rawi configure --profile creative \
  --provider openai \
  --model gpt-4 \
  --temperature 0.9 \
  --max-tokens 3000

# Technical documentation profile
rawi configure --profile docs \
  --provider anthropic \
  --model claude-3-5-sonnet-20241022 \
  --temperature 0.3 \
  --max-tokens 2500

Using Profiles

# Use specific profile
rawi ask "Review this code" --profile code-review

# List all profiles
rawi configure --list

# Show profile details
rawi configure --profile work --show

# Set default profile
rawi configure --profile work --set-default

Provider-Specific Configuration

OpenAI Configuration

# Basic OpenAI setup
rawi configure --provider openai --model gpt-4o --api-key sk-your-key

# Advanced OpenAI setup
rawi configure --provider openai \
  --model gpt-4o \
  --api-key sk-your-key \
  --temperature 0.7 \
  --max-tokens 2048 \
  --base-url https://api.openai.com/v1

Anthropic Configuration

# Claude setup
rawi configure --provider anthropic \
  --model claude-3-5-sonnet-20241022 \
  --api-key sk-ant-your-key \
  --temperature 0.1 \
  --max-tokens 4000

Azure OpenAI Configuration

# Azure OpenAI setup
rawi configure --provider azure \
  --model gpt-4 \
  --api-key your-azure-key \
  --resource-name your-resource \
  --base-url https://your-resource.openai.azure.com \
  --api-version 2024-10-01-preview

Local Providers (Ollama)

# Ollama setup
rawi configure --provider ollama \
  --model llama2 \
  --base-url http://localhost:11434

# Custom Ollama model
rawi configure --provider ollama \
  --model codellama:13b \
  --base-url http://localhost:11434

Environment Variables

You can configure Rawi using environment variables for automation and CI/CD:

API Keys

# Common API keys
export OPENAI_API_KEY="sk-your-openai-key"
export ANTHROPIC_API_KEY="sk-ant-your-anthropic-key"
export GOOGLE_API_KEY="your-google-key"

# Azure-specific
export AZURE_OPENAI_API_KEY="your-azure-key"
export AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com"

# AWS Bedrock
export AWS_ACCESS_KEY_ID="your-access-key"
export AWS_SECRET_ACCESS_KEY="your-secret-key"
export AWS_REGION="us-east-1"

Default Settings

# Default profile and provider
export RAWI_PROFILE="work"
export RAWI_PROVIDER="openai"
export RAWI_MODEL="gpt-4"

# Generation settings
export RAWI_TEMPERATURE="0.7"
export RAWI_MAX_TOKENS="2048"
export RAWI_LANGUAGE="english"

Configuration Files

Rawi stores configuration in structured files:

Directory Structure

~/.config/rawi/
├── config.json          # Global settings
├── profiles/
│   ├── default.json      # Default profile
│   ├── work.json         # Work profile
│   └── personal.json     # Personal profile
├── history/              # Session history
├── templates/            # Custom templates
└── cache/                # Temporary cache

Profile File Example

{
  "name": "work",
  "provider": "azure",
  "model": "gpt-4",
  "apiKey": "your-azure-key",
  "baseUrl": "https://company-openai.openai.azure.com",
  "resourceName": "company-openai",
  "temperature": 0.3,
  "maxTokens": 2048,
  "language": "english",
  "createdAt": "2024-01-15T10:30:00Z",
  "lastUsed": "2024-01-20T15:45:00Z"
}

Advanced Configuration

Custom Base URLs

# Custom OpenAI endpoint
rawi configure --provider openai \
  --base-url https://custom-openai-proxy.com/v1 \
  --api-key your-key

# Local LM Studio
rawi configure --provider openai \
  --base-url http://localhost:1234/v1 \
  --model local-model

Regional Configuration

# AWS Bedrock in different region
rawi configure --provider bedrock \
  --region eu-west-1 \
  --model anthropic.claude-3-sonnet-20240229-v1:0

# Azure in specific region
rawi configure --provider azure \
  --model gpt-4 \
  --base-url https://westeurope-openai.openai.azure.com

Proxy Configuration

# HTTP proxy
export HTTP_PROXY="http://proxy.company.com:8080"
export HTTPS_PROXY="http://proxy.company.com:8080"

# Custom CA bundle
rawi configure --ca-bundle /path/to/custom-ca.pem

Configuration Best Practices

Security

Use Environment Variables: Store API keys in environment variables for production
Profile Separation: Use different profiles for different security contexts
Regular Rotation: Rotate API keys regularly
Local Providers: Use Ollama for sensitive data processing

Performance

Model Selection: Choose appropriate models for your use case
Token Limits: Set reasonable token limits to control costs
Temperature Settings: Lower for factual tasks, higher for creative work
Profile Optimization: Create task-specific profiles

Organization

Descriptive Names: Use clear profile names like work-gpt4, personal-claude
Environment Profiles: Separate dev, staging, and production configurations
Team Standards: Standardize configurations across team members
Documentation: Document your profile configurations

Troubleshooting Configuration

Common Issues

Configuration not saved:

# Check permissions
ls -la ~/.config/rawi/
# Create directory if missing
mkdir -p ~/.config/rawi/profiles/

API key not working:

# Test configuration
rawi configure --test
# Verify API key format
rawi configure --show

Provider connection fails:

# Check network connectivity
curl -s https://api.openai.com/v1/models -H "Authorization: Bearer $OPENAI_API_KEY"
# Try different provider
rawi configure --provider anthropic

Configuration Validation

# Test current configuration
rawi ask "test" --verbose

# Validate specific profile
rawi configure --profile work --test

# Check provider status
rawi info --providers

Migration and Backup

Export Configuration

# Export all profiles
rawi configure --export > rawi-config-backup.json

# Backup configuration directory
cp -r ~/.config/rawi/ ~/rawi-backup-$(date +%Y%m%d)/

Import Configuration

# Import configuration
rawi configure --import rawi-config-backup.json

# Restore from backup
cp -r ~/rawi-backup-20240115/ ~/.config/rawi/

Next Steps

After configuring Rawi:

Learn about specific providers: Supported Providers
Explore environment variables: Environment Variables
Set up act templates: Act Templates
Configure advanced options: Advanced Settings
Start using Rawi: Quick Start Guide