`OSSS.ai.config.agent_configs`¶

`OSSS.ai.config.agent_configs` ¶

Agent Configuration Classes for Configurable Prompt Composition

This module provides Pydantic-based configuration classes for dynamic agent behavior modification. Each agent can be configured via YAML workflows or environment variables while maintaining full backward compatibility.

Architecture: - Base configuration classes for common patterns - Agent-specific configurations with validation - Integration with existing prompt system - Environment variable and workflow loading

`PromptConfig` ¶

Bases: BaseModel

Base configuration for prompt-related settings.

`BehavioralConfig` ¶

Bases: BaseModel

Base configuration for agent behavioral patterns.

`OutputConfig` ¶

Bases: BaseModel

Base configuration for output formatting and structure.

`AgentExecutionConfig` ¶

Bases: BaseModel

Base configuration for execution behavior.

`RefinerConfig` ¶

Bases: BaseModel

Configuration for RefinerAgent behavior and prompt composition.

`from_dict(config)` `classmethod` ¶

Create RefinerConfig from dictionary (workflow integration).

`from_env(prefix='REFINER')` `classmethod` ¶

Create RefinerConfig from environment variables.

`to_prompt_config()` ¶

Convert to format compatible with existing prompt system.

`CriticConfig` ¶

Bases: BaseModel

Configuration for CriticAgent behavior and prompt composition.

`from_dict(config)` `classmethod` ¶

Create CriticConfig from dictionary (workflow integration).

`from_env(prefix='CRITIC')` `classmethod` ¶

Create CriticConfig from environment variables.

`to_prompt_config()` ¶

Convert to format compatible with existing prompt system.

`HistorianConfig` ¶

Bases: BaseModel

Configuration for HistorianAgent behavior and prompt composition.

`from_dict(config)` `classmethod` ¶

Create HistorianConfig from dictionary (workflow integration).

`from_env(prefix='HISTORIAN')` `classmethod` ¶

Create HistorianConfig from environment variables.

`to_prompt_config()` ¶

Convert to format compatible with existing prompt system.

`SynthesisConfig` ¶

Bases: BaseModel

Configuration for SynthesisAgent behavior and prompt composition.

`from_dict(config)` `classmethod` ¶

Create SynthesisConfig from dictionary (workflow integration).

`from_env(prefix='SYNTHESIS')` `classmethod` ¶

Create SynthesisConfig from environment variables.

`to_prompt_config()` ¶

Convert to format compatible with existing prompt system.

`FinalConfig` ¶

Bases: BaseModel

Configuration for FinalAgent behavior and prompt composition.

This aligns with FinalAgent's responsibilities: - Build the final user-facing answer - Integrate RAG context when present - Enforce strict role-identity guardrails

`from_dict(config)` `classmethod` ¶

Create FinalConfig from dictionary (workflow integration).

`from_env(prefix='FINAL')` `classmethod` ¶

Create FinalConfig from environment variables.

`to_prompt_config()` ¶

Convert to a flat dict compatible with the existing prompt system.

FinalAgent can read these from ctx.execution_state["execution_config"] or a similar structure.

`get_agent_config_class(agent_type)` ¶

get_agent_config_class(
    agent_type: Literal["refiner"],
) -> Type[RefinerConfig]

get_agent_config_class(
    agent_type: Literal["critic"],
) -> Type[CriticConfig]

get_agent_config_class(
    agent_type: Literal["historian"],
) -> Type[HistorianConfig]

get_agent_config_class(
    agent_type: Literal["synthesis"],
) -> Type[SynthesisConfig]

get_agent_config_class(
    agent_type: Literal["final"],
) -> Type[FinalConfig]

get_agent_config_class(
    agent_type: str,
) -> Type[AgentConfigType]

Get the appropriate configuration class for an agent type.

`create_agent_config(agent_type, config_dict)` ¶

create_agent_config(
    agent_type: Literal["refiner"],
    config_dict: Dict[str, Any],
) -> RefinerConfig

create_agent_config(
    agent_type: Literal["critic"],
    config_dict: Dict[str, Any],
) -> CriticConfig

create_agent_config(
    agent_type: Literal["historian"],
    config_dict: Dict[str, Any],
) -> HistorianConfig

create_agent_config(
    agent_type: Literal["synthesis"],
    config_dict: Dict[str, Any],
) -> SynthesisConfig

create_agent_config(
    agent_type: Literal["final"],
    config_dict: Dict[str, Any],
) -> FinalConfig

create_agent_config(
    agent_type: str, config_dict: Dict[str, Any]
) -> AgentConfigType

Factory function to create appropriate agent configuration from dictionary.

OSSS.ai.config.agent_configs¶