agents.simple.enhanced_simple_real¶
Enhanced SimpleAgent - Real implementation using Agent[AugLLMConfig].
This is the real SimpleAgent implementation showing it as Agent[AugLLMConfig]. It carefully imports only what’s needed to avoid circular imports.
Classes¶
Minimal base for enhanced agents to avoid circular imports. |
|
Enhanced SimpleAgent that is essentially Agent[AugLLMConfig]. |
Module Contents¶
- class agents.simple.enhanced_simple_real.EnhancedAgentBase¶
Minimal base for enhanced agents to avoid circular imports.
- async arun(input_data)¶
Async run method.
- Parameters:
input_data (Any)
- Return type:
Any
- abstractmethod build_graph()¶
Build the agent’s graph.
- Return type:
Any
- run(input_data)¶
Sync run method.
- Parameters:
input_data (Any)
- Return type:
Any
- setup_agent()¶
Hook for subclass setup.
- Return type:
None
- class agents.simple.enhanced_simple_real.SimpleAgent¶
Bases:
EnhancedAgentBase
Enhanced SimpleAgent that is essentially Agent[AugLLMConfig].
This demonstrates the key insight: SimpleAgent IS Agent[AugLLMConfig]. All the complexity is handled by the base Agent class and the engine type.
In the full implementation with working imports, this would inherit from: Agent[AugLLMConfig] where Agent is from enhanced_agent.py
Key points: - Engine is always AugLLMConfig - Minimal implementation needed - Type safety for engine-specific features - Clean separation of concerns
- build_graph()¶
Build minimal graph for SimpleAgent.
- Return type:
haive.core.graph.state_graph.base_graph2.BaseGraph
- classmethod ensure_aug_llm_config(values)¶
Ensure we have an AugLLMConfig engine.
- setup_agent()¶
Sync convenience fields to the AugLLMConfig engine.
- Return type:
None