kg6-codex
Core Concepts

Configuration

Learn how to configure AI Developer Assistant for your needs

Configuration

AI Developer Assistant can be configured through multiple methods: environment variables, configuration files, and command-line options. This guide covers all configuration options and best practices.

Configuration Methods

1. Environment Variables (Recommended)

Set environment variables for persistent configuration:

# Add to your shell profile (.bashrc, .zshrc, etc.)
export LLM_PROVIDER="gemini"
export GEMINI_API_KEY="your-gemini-api-key"
export LLM_MODEL="gemini-2.0-flash"
export LLM_TEMPERATURE="0.7"
export LLM_MAX_TOKENS="2000"
export OUTPUT_FORMAT="console"
export OUTPUT_VERBOSE="true"
export OUTPUT_COLORIZE="true"

2. Configuration Files

Create configuration files for project-specific settings:

# Copy default configuration
cp ai-dev.config.yaml ai-dev.config.local.yaml

# Edit with your settings
nano ai-dev.config.local.yaml

3. Command-Line Options

Override configuration for specific commands:

ai-dev review --llm-provider openai --openai-api-key your-key --verbose

Configuration Priority

Configuration is applied in this order (later overrides earlier):

  1. Default values (built-in)
  2. Configuration files (ai-dev.config.local.yaml)
  3. Environment variables
  4. Command-line options

LLM Provider Configuration

OpenAI Configuration

llm:
  provider: "openai"
  model: "gpt-4"
  temperature: 0.7
  maxTokens: 2000

openai:
  apiKey: "${OPENAI_API_KEY}"
  baseUrl: "https://api.openai.com/v1"
  organization: "${OPENAI_ORGANIZATION}"  # Optional

Environment Variables:

export OPENAI_API_KEY="sk-your-openai-key"
export OPENAI_ORGANIZATION="org-your-org-id"  # Optional
export LLM_PROVIDER="openai"
export LLM_MODEL="gpt-4"

Google Gemini Configuration

llm:
  provider: "gemini"
  model: "gemini-2.0-flash"
  temperature: 0.7
  maxTokens: 2000

gemini:
  apiKey: "${GEMINI_API_KEY}"
  model: "gemini-2.0-flash"
  baseUrl: "https://generativelanguage.googleapis.com/v1beta"

Environment Variables:

export GEMINI_API_KEY="AIzaSy-your-gemini-key"
export LLM_PROVIDER="gemini"
export LLM_MODEL="gemini-2.0-flash"

Ollama (Local) Configuration

llm:
  provider: "ollama"
  model: "llama2"
  temperature: 0.7
  maxTokens: 2000

ollama:
  enabled: true
  baseUrl: "http://localhost:11434"
  model: "llama2"

Environment Variables:

export OLLAMA_ENABLED="true"
export OLLAMA_BASE_URL="http://localhost:11434"
export OLLAMA_MODEL="llama2"
export LLM_PROVIDER="ollama"

Output Configuration

Console Output

output:
  format: "console"
  colorize: true
  verbose: false

Markdown Output

output:
  format: "markdown"
  outputPath: "./reports/review.md"
  colorize: false
  verbose: true

JSON Output

output:
  format: "json"
  outputPath: "./reports/review.json"
  colorize: false
  verbose: false

Security Configuration

Basic Security Settings

security:
  enabled: true
  severity: ["medium", "high", "critical"]
  categories: ["injection", "authentication", "authorization"]
  includeDependencies: true

Advanced Security Settings

security:
  enabled: true
  severity: ["low", "medium", "high", "critical"]
  categories:
    - "injection"
    - "authentication"
    - "authorization"
    - "cryptography"
    - "data_exposure"
    - "input_validation"
    - "dependency"
    - "configuration"
    - "logging"
    - "other"
  includeDependencies: true
  customRules: []

Test Configuration

Basic Test Settings

test:
  enabled: true
  framework: "jest"
  testType: "unit"
  language: "typescript"
  coverageTarget: 80

Advanced Test Settings

test:
  enabled: true
  framework: "jest"
  testType: "unit"
  language: "typescript"
  includeSetup: true
  includeTeardown: true
  coverageTarget: 80
  customTemplates: []

Documentation Configuration

Basic Documentation Settings

documentation:
  enabled: true
  format: "markdown"
  language: "typescript"
  includeExamples: true

Advanced Documentation Settings

documentation:
  enabled: true
  format: "markdown"
  language: "typescript"
  includeExamples: true
  includeParameters: true
  includeReturnTypes: true
  includeSeeAlso: true
  style: "technical"  # formal, casual, technical, beginner

Git Configuration

Basic Git Settings

git:
  repoPath: "."
  defaultBranch: "main"
  includeStaged: true
  includeUnstaged: true

Advanced Git Settings

git:
  repoPath: "."
  defaultBranch: "main"
  includeStaged: true
  includeUnstaged: true
  diffAlgorithm: "histogram"  # myers, minimal, patience, histogram
  ignoreWhitespace: false
  maxFileSize: "1MB"

GitHub Integration Configuration

Basic GitHub Settings

github:
  token: "${GITHUB_TOKEN}"
  baseUrl: "https://api.github.com"
  apiVersion: "2022-11-28"

Advanced GitHub Settings

github:
  token: "${GITHUB_TOKEN}"
  baseUrl: "https://api.github.com"
  apiVersion: "2022-11-28"
  timeout: 30000
  retries: 3
  rateLimitRetries: 3

File Pattern Configuration

Include Patterns

filePatterns:
  - "**/*.ts"
  - "**/*.js"
  - "**/*.tsx"
  - "**/*.jsx"
  - "**/*.py"
  - "**/*.java"
  - "**/*.cpp"
  - "**/*.c"
  - "**/*.cs"
  - "**/*.php"
  - "**/*.rb"
  - "**/*.go"
  - "**/*.rs"
  - "**/*.dart"

Exclude Patterns

excludePatterns:
  - "**/node_modules/**"
  - "**/dist/**"
  - "**/build/**"
  - "**/.git/**"
  - "**/coverage/**"
  - "**/*.min.js"
  - "**/*.bundle.js"

Complete Configuration Example

Here's a complete configuration file with all options:

# AI Developer Assistant Configuration
llm:
  provider: "gemini"
  model: "gemini-2.0-flash"
  temperature: 0.7
  maxTokens: 2000
  timeout: 30000

# LLM Provider Configurations
openai:
  apiKey: "${OPENAI_API_KEY}"
  baseUrl: "https://api.openai.com/v1"
  organization: "${OPENAI_ORGANIZATION}"

ollama:
  enabled: false
  baseUrl: "http://localhost:11434"
  model: "llama2"

gemini:
  apiKey: "${GEMINI_API_KEY}"
  model: "gemini-2.0-flash"
  baseUrl: "https://generativelanguage.googleapis.com/v1beta"

# GitHub Integration
github:
  token: "${GITHUB_TOKEN}"
  baseUrl: "https://api.github.com"
  apiVersion: "2022-11-28"

# Git Configuration
git:
  repoPath: "."
  defaultBranch: "main"
  includeStaged: true
  includeUnstaged: true

# Output Configuration
output:
  format: "console"
  colorize: true
  verbose: false

# Security Scanning
security:
  enabled: true
  severity: ["medium", "high", "critical"]
  categories: ["injection", "authentication", "authorization"]
  includeDependencies: true

# Test Generation
test:
  enabled: true
  framework: "jest"
  testType: "unit"
  language: "typescript"
  coverageTarget: 80

# Documentation Generation
documentation:
  enabled: true
  format: "markdown"
  language: "typescript"
  includeExamples: true

# File Patterns
filePatterns:
  - "**/*.ts"
  - "**/*.js"
  - "**/*.py"
  - "**/*.java"

excludePatterns:
  - "**/node_modules/**"
  - "**/dist/**"
  - "**/build/**"

# Global Settings
verbose: false
debug: false

Configuration Validation

Validate Configuration

# Validate your configuration
ai-dev config validate

# Show current configuration
ai-dev config show

# Test configuration with specific provider
ai-dev config test --provider gemini

Common Configuration Issues

Invalid YAML syntax:

# Check for syntax errors
ai-dev config validate

# Common issues:
# - Incorrect indentation (use 2 spaces)
# - Missing quotes around values with special characters
# - Invalid boolean values (use true/false, not True/False)

Missing required fields:

# Check for missing API keys
ai-dev config show | grep -E "(apiKey|token)"

# Set missing values
export GEMINI_API_KEY="your-key"

Best Practices

Security

  1. Never commit API keys:
# Add to .gitignore
echo "ai-dev.config.local.yaml" >> .gitignore
echo ".env" >> .gitignore
  1. Use environment variables:
# Instead of hardcoding
apiKey: "sk-your-key"

# Use environment variables
apiKey: "${OPENAI_API_KEY}"

Organization

  1. Use project-specific configs:
# Project-specific configuration
cp ai-dev.config.yaml ai-dev.config.local.yaml
  1. Separate concerns:
# Different configs for different purposes
ai-dev.config.dev.yaml    # Development
ai-dev.config.prod.yaml   # Production
ai-dev.config.test.yaml   # Testing

Performance

  1. Optimize file patterns:
# Be specific with patterns
filePatterns:
  - "src/**/*.ts"           # Instead of "**/*.ts"
  - "lib/**/*.dart"         # Instead of "**/*.dart"
  1. Exclude unnecessary files:
excludePatterns:
  - "**/node_modules/**"
  - "**/dist/**"
  - "**/*.min.js"
  - "**/coverage/**"
Start with basic configuration and gradually add advanced features as needed. The default configuration works well for most use cases.

Usage Patterns & Workflows

Learn how to use AI Developer Assistant effectively in your development workflow

Output Formatting & Themes

Learn how to customize AI Developer Assistant output formatting and themes