Configuration

The config section key is ai (configures AIConfig). Environment prefix: LEX_AI__.

AIConfig

Key	Type	Default	Env Var	Description
`name`	`str`	`"ai"`	`LEX_AI__NAME`	Configuration name
`enabled`	`bool`	`True`	`LEX_AI__ENABLED`	Enable AI features
`llm`	`ClientConfig \| None`	`None`	`LEX_AI__LLM__*`	LLM client configuration
`vector`	`VectorConfig \| None`	`None`	`LEX_AI__VECTOR__*`	Vector store configuration
`rag`	`RAGConfig \| None`	`None`	`LEX_AI__RAG__*`	RAG pipeline configuration
`governance`	`GovernanceConfig`	default-factory (disabled when governance package absent)	`LEX_AI__GOVERNANCE__*`	AI governance configuration
`observability`	`ObservabilityConfig`	default-factory (disabled if `lexigram-ai-observability` not installed)	`LEX_AI__OBSERVABILITY__*`	AI observability configuration
`subsystems`	`dict[str, dict[str, Any]]`	`{}`	`LEX_AI__SUBSYSTEMS__*`	Dynamic config for third-party subsystems

Example YAML

ai:
  enabled: true
  llm:
    provider: openai
    model: gpt-4o
    api_key: ${LEX_AI_LLM__API_KEY}
    temperature: 0.7
    timeout: 60
  governance:
    enabled: true
  subsystems:
    custom_subsystem:
      option_a: value

Env-var overrides

The nested delimiter is __ (double underscore):

export LEX_AI__ENABLED=true
export LEX_AI_LLM__PROVIDER=anthropic
export LEX_AI_LLM__MODEL=claude-sonnet-4-20250514
export LEX_AI_LLM__API_KEY=sk-ant-...
export LEX_AI_LLM__TEMPERATURE=0.3

Subsystem configuration

Use get_subsystem_config() to retrieve config for a dynamically-discovered subsystem:

from lexigram.ai.config import AIConfig, get_subsystem_config

config = AIConfig()
sub_cfg = get_subsystem_config(config, "llm")
# Returns config.llm if set, otherwise config.subsystems["llm"]

Production security

AIConfig includes a model_validator that blocks insecure API keys (sk-..., sk-test, change-me) when LEX_ENV=production.