agents.planning.llm_compiler_v3.state¶

State schema for LLM Compiler V3 Agent.

Classes¶

State schema for LLM Compiler V3 Agent using Enhanced MultiAgent V3.

class agents.planning.llm_compiler_v3.state.LLMCompilerStateSchema(/, **data)¶

Bases: haive.core.schema.prebuilt.messages_state.MessagesState

State schema for LLM Compiler V3 Agent using Enhanced MultiAgent V3.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

add_execution_result(result)¶

Add a task execution result to state.

Parameters:: result (haive.agents.planning.llm_compiler_v3.models.ParallelExecutionResult)
Return type:: None

can_execute_more_tasks()¶

Check if more tasks can be executed in parallel.

get_execution_summary()¶

Get comprehensive execution summary.

get_failed_results()¶

Get all failed execution results.

Return type:: list[haive.agents.planning.llm_compiler_v3.models.ParallelExecutionResult]

get_next_executable_tasks(count=None)¶

Get the next tasks to execute, respecting parallel limits.

Parameters:: count (int | None)
Return type:: list[haive.agents.planning.llm_compiler_v3.models.CompilerTask]

get_successful_results()¶

Get all successful execution results.

Return type:: list[haive.agents.planning.llm_compiler_v3.models.ParallelExecutionResult]

is_execution_complete()¶

Check if all tasks in the plan have been executed or failed.

mark_task_executing(task_id)¶

Mark a task as currently executing.

resolve_task_arguments(task)¶

Resolve task arguments by substituting dependency references.

Parameters:: task (haive.agents.planning.llm_compiler_v3.models.CompilerTask)
Return type:: dict[str, Any]

should_replan()¶

Determine if replanning is needed based on execution state.

update_ready_tasks()¶

Update lists of ready and blocked tasks based on current state.