Aurelis

Getting Started with Aurelis

Aurelis is an enterprise-grade AI code assistant that exclusively uses GitHub models via Azure AI Inference. This guide will get you up and running in minutes.

🚀 Quick Start

Prerequisites

• Python 3.8+ installed on your system
• GitHub Account with access to GitHub models
• GitHub Token with appropriate model access permissions

1. Installation

# Install from PyPI (recommended)
pip install aurelis-cli

# Or install from source
git clone https://github.com/kanopusdev/aurelis.git
cd aurelis
pip install -e .

2. GitHub Token Setup

Aurelis uses GitHub models exclusively through Azure AI Inference. You need a GitHub token with model access.

Get Your GitHub Token

1. Go to GitHub Settings → Personal Access Tokens
2. Click “Generate new token (classic)”
3. Select appropriate scopes for model access
4. Copy the generated token

Set Environment Variable

# Linux/macOS
export GITHUB_TOKEN="your_github_token_here"

# Windows PowerShell
$env:GITHUB_TOKEN="your_github_token_here"

# Windows Command Prompt
set GITHUB_TOKEN=your_github_token_here

3. Initialize Aurelis

# Initialize configuration
aurelis init

# Verify setup
aurelis models

4. First Commands

# List available GitHub models
aurelis models

# Analyze a Python file
aurelis analyze script.py

# Generate code from description
aurelis generate "Create a FastAPI endpoint for user authentication"

# Start interactive shell
aurelis shell

🖥️ Interactive Shell

The Aurelis shell provides the most powerful experience:

aurelis shell

Inside the shell, you can use these commands:

# Model management
models          # List available models
health          # Check model connectivity
config          # View configuration

# Code operations
analyze file.py         # Analyze code
generate "description"  # Generate code
explain "code snippet"  # Explain code
fix file.py            # Fix issues
refactor file.py       # Optimize code

# Documentation
docs file.py           # Generate docs
test file.py           # Create tests

# Session management
session save name      # Save session
session load name      # Load session
history               # Command history
help                  # Show all commands
exit                  # Exit shell

📁 Project Configuration

Create a .aurelis.yaml file in your project root:

# GitHub Models Configuration
github_token: "${GITHUB_TOKEN}"  # Use environment variable

models:
  primary: "codestral-2501"       # Primary model for code tasks
  fallback: "gpt-4o-mini"         # Fallback model for reliability
  
analysis:
  max_file_size: "1MB"
  chunk_size: 3500               # Optimized for 4K context models
  overlap_ratio: 0.15
  
processing:
  max_retries: 3
  timeout: 60
  concurrent_requests: 5
  
security:
  audit_logging: true
  secure_token_storage: true

cache:
  enabled: true
  ttl: 3600  # 1 hour
  max_size: 1000

🤖 Understanding GitHub Models

Aurelis intelligently routes tasks to the best GitHub model:

Code Generation Tasks

• Primary: Codestral-2501 (specialized for coding)
• Fallback: GPT-4o (complex reasoning)

Documentation Tasks

• Primary: Cohere Command-R (optimized for explanations)
• Fallback: GPT-4o-mini (fast responses)

Complex Reasoning

• Primary: GPT-4o (advanced reasoning)
• Fallback: Meta Llama 3.1 405B (maximum capability)

Performance Tasks

• Primary: Mistral Nemo (fast inference)
• Fallback: GPT-4o-mini (efficient responses)

✅ Verification

Test your setup with these commands:

# Check configuration
aurelis config

# Verify token and connectivity
aurelis models

# Test code analysis
echo "def hello(): print('world')" > test.py
aurelis analyze test.py

# Test code generation
aurelis generate "Python function to calculate fibonacci"

# Test shell
aurelis shell

🐛 Troubleshooting

Common Issues

“GitHub token not configured”

# Verify token is set
echo $GITHUB_TOKEN  # Linux/macOS
echo $env:GITHUB_TOKEN  # Windows PowerShell

# Set token if missing
export GITHUB_TOKEN="your_token_here"

“Failed to initialize Aurelis”

# Check configuration
aurelis config

# Recreate configuration
aurelis init --force

“Model not responding”

# Check model health
aurelis models

# Verify network connectivity
curl -H "Authorization: Bearer $GITHUB_TOKEN" \
  https://models.inference.ai.azure.com/health

“Import errors”

# Reinstall dependencies
pip install --upgrade aurelis

# Or install from source
pip install -e .

Getting Help

# CLI help
aurelis --help

# Command-specific help
aurelis analyze --help
aurelis generate --help

# Shell help
aurelis shell
> help

🏢 Enterprise Setup

For enterprise setups, see:

• Enterprise Architecture
• Security Configuration
• Performance Configuration

📖 Next Steps

Now that you have Aurelis running:

1. Explore Commands: Try all CLI commands with your code
2. Master the Shell: Use the interactive shell for complex tasks
3. Configure Models: Customize model preferences for your workflow
4. Integrate in IDE: Set up IDE integration for seamless development
5. Team Setup: Configure for team collaboration

Learning Resources

• User Guide - Comprehensive usage documentation
• CLI Reference - Complete command reference
• GitHub Models Guide - Model selection and optimization
• Best Practices - Production tips and patterns

---

Need Help?

• 📖 Documentation
• 🐛 Report Issues
• 💬 Discussions
• 🏢 Enterprise Support

Ready to supercharge your development with AI? Let’s code! 🚀