Configuration Guide

Environment Variables

Create a .env file in your project root or set environment variables.

⚠️ SECURITY WARNING: Always add .env to your .gitignore file to prevent accidentally committing API keys (GEMINI_API_KEY, OPENAI_API_KEY, ANTHROPIC_API_KEY) to version control. Exposed API keys can lead to unauthorized usage and significant costs.
echo ".env" >> .gitignore

Type Reference

Type	Description	Examples
`string`	Text value	`gemini-2.0-flash-exp`, `docs/analyses`
`integer`	Whole number (constraints noted in parentheses)	`10`, `3600`
`boolean`	True/false flag	`true`, `false`
`enum`	Closed set of valid values (listed in description)	`gemini`, `medium`

Core Settings

Variable	Type	Description	Default
`AI_PROVIDER`	`enum`	AI provider. Valid values: `gemini`, `openai`, `claude`	`gemini`
`AI_MODEL`	`string`	Model identifier (provider-specific)	Provider default
`GEMINI_API_KEY`	`string`	Gemini API key	-
`OPENAI_API_KEY`	`string`	OpenAI API key	-
`ANTHROPIC_API_KEY`	`string`	Claude API key	-

Analysis Settings

Variable	Type	Description	Default
`ROOT_DIR`	`string`	Directory path to analyze	`.`
`OUTPUT_DIR`	`string`	Output directory path for reports	`docs/analyses`
`ARCHITECTURE_REPORT`	`string`	Architecture report file path	`docs/architecture_overview.md`
`BATCH_SIZE`	`integer`	Number of files analyzed in parallel (≥1)	`10`
`MAX_RETRIES`	`integer`	Retry attempts on error (≥0)	`3`
`TIMEOUT_SECONDS`	`integer`	Request timeout in seconds (≥1)	`60`

Feature Toggles

Variable	Type	Description	Default
`ENABLE_WEB_SEARCH`	`boolean`	Enable documentation search	`true`
`ENABLE_SECURITY_ANALYSIS`	`boolean`	Enable security analysis	`true`
`ENABLE_METRICS_ANALYSIS`	`boolean`	Enable code metrics	`true`
`ENABLE_DEPENDENCY_VULNERABILITY`	`boolean`	Enable CVE scanning	`true`
`ENABLE_CODE_GENERATION`	`boolean`	Enable code analysis generation	`true`
`SHOW_REASONING`	`boolean`	Show AI reasoning process	`false`

Cache Settings

Variable	Type	Description	Default
`ENABLE_CACHE`	`boolean`	Enable analysis caching	`true`
`CACHE_TTL_SECONDS`	`integer`	Cache lifetime in seconds (≥0)	`604800` (7 days)

Note: All duration settings use seconds as the unit for consistency. Common values: 1 hour = 3600, 1 day = 86400, 7 days = 604800.

Dependency Analysis

Variable	Type	Description	Default
`ANALYZE_DEPENDENT_FILES`	`boolean`	Analyze dependent files	`true`
`MAX_DEPENDENCY_DEPTH`	`integer`	Max dependency depth (≥0)	`2`
`ENABLE_EXPORT_ANALYSIS`	`boolean`	Extract exports	`true`
`ENABLE_REVERSE_DEPENDENCY`	`boolean`	Reverse dependency graph	`true`
`ENABLE_IMPACT_ANALYSIS`	`boolean`	Bug impact analysis	`true`
`MAX_IMPACT_DEPTH`	`integer`	Max impact depth (≥0)	`3`

Gemini Advanced Features

Variable	Type	Description	Default
`ENABLE_STRUCTURED_OUTPUTS`	`boolean`	JSON validated via Pydantic	`true`
`ENABLE_THINKING`	`boolean`	Show reasoning process	`false`
`THINKING_BUDGET`	`integer`	Thinking token budget (≥0)	`1024`
`ENABLE_CODE_EXECUTION`	`boolean`	Python sandbox execution	`false`
`ENABLE_FUNCTION_CALLING`	`boolean`	Auto function calling	`false`
`ENABLE_CONTEXT_CACHING`	`boolean`	Context cache (-50% cost)	`true`
`CONTEXT_CACHE_TTL_SECONDS`	`integer`	Context cache TTL in seconds (≥0)	`3600`
`ENABLE_BATCH_MODE`	`boolean`	Batch API (-50% cost)	`false`

Bug Fixing

Variable	Type	Description	Default
`ENABLE_CLAUDE_CODE_FIX`	`boolean`	Enable Claude Code integration	`true`
`FIX_SEVERITY_THRESHOLD`	`enum`	Minimum severity to fix. Valid values: `low`, `medium`, `high`, `critical`	`medium`

Extended Agents (v2.10)

Variable	Type	Description	Default
`ENABLE_TYPE_ANALYSIS`	`boolean`	Enable Type Safety Agent (any usage, unsafe casts)	`true`
`ENABLE_PERFORMANCE_ANALYSIS`	`boolean`	Enable Performance Agent (O(n²), N+1 queries)	`true`
`ENABLE_API_CONTRACT_ANALYSIS`	`boolean`	Enable API Contract Agent (REST violations)	`true`
`ENABLE_PRIVACY_ANALYSIS`	`boolean`	Enable Data Privacy Agent (PII, compliance)	`true`

Agent Selection (v2.10)

Control which agents run during analysis:

# Run specific agents only
aetheris analysis --agents security,performance,type_safety

# Skip specific agents
aetheris analysis --skip-agents metrics,architecture

# Enable compliance frameworks for privacy analysis
aetheris analysis --compliance gdpr,hipaa,pci_dss

Available agents: - Core: code_analysis, architecture, security, metrics, vulnerability, qa - Extended: type_safety, performance, api_contract, data_privacy

Compliance frameworks: - gdpr — EU General Data Protection Regulation - hipaa — US Health Insurance Portability Act - pci_dss — Payment Card Industry Data Security Standard - ccpa — California Consumer Privacy Act

Example `.env` File

# AI Provider
AI_PROVIDER=gemini
GEMINI_API_KEY=   # Set via environment variable or secrets manager
AI_MODEL=gemini-2.0-flash-exp

# Analysis
BATCH_SIZE=10
TIMEOUT_SECONDS=120
OUTPUT_DIR=docs/analyses

# Features
ENABLE_SECURITY_ANALYSIS=true
ENABLE_METRICS_ANALYSIS=true
ENABLE_DEPENDENCY_VULNERABILITY=true

# Cache
ENABLE_CACHE=true
CACHE_TTL_SECONDS=604800  # 7 days

# Gemini Advanced
ENABLE_STRUCTURED_OUTPUTS=true
ENABLE_CONTEXT_CACHING=true

Security Note: Never commit API keys to version control. Use a .env file (added to .gitignore) or a secrets manager like HashiCorp Vault, AWS Secrets Manager, or Azure Key Vault.

Setting Environment Variables

Linux/macOS

export AI_PROVIDER=gemini
export GEMINI_API_KEY="$GEMINI_API_KEY"  # Reference from secure source
# Or load from a secrets manager:
# export GEMINI_API_KEY=$(vault kv get -field=api_key secret/aetheris)

Windows (PowerShell)

$env:AI_PROVIDER = 'gemini'
$env:GEMINI_API_KEY = $env:GEMINI_API_KEY  # Reference from secure source
# Or use PowerShell SecretManagement:
# $env:GEMINI_API_KEY = Get-Secret -Name 'GeminiApiKey' -AsPlainText

Windows (CMD)

set AI_PROVIDER=gemini
set GEMINI_API_KEY=%GEMINI_API_KEY%
rem Reference from another secure source or set interactively

Next Steps

AI Providers - Models and pricing
Commands Reference - Available commands