src.core.ai_powered_task_assignment module#

AI-Powered Task Assignment for Marcus.

This module upgrades the basic task assignment logic to use Phase 1-4 AI capabilities for intelligent task selection.

class src.core.ai_powered_task_assignment.AITaskAssignmentEngine[source]#

Bases: object

Intelligent task assignment using Phase 1-4 capabilities.

Features: - Phase 1: Safety checks (no deploy before implement) - Phase 2: Dependency analysis (prioritize unblocking tasks) - Phase 3: AI-powered agent matching - Phase 4: Predictive impact analysis

__init__(ai_engine, project_tasks, ai_analysis_engine=None)[source]#

Parameters:

ai_engine (MarcusAIEngine)
project_tasks (List[Task])
ai_analysis_engine (AIAnalysisEngine | None)

async find_optimal_task_for_agent(agent_id, agent_info, available_tasks, assigned_task_ids)[source]#

Find the best task for an agent using AI capabilities.

This replaces the basic skill/priority matching with intelligent analysis.

Return type:

Optional[Task]

Parameters:

agent_id (str)
agent_info (Dict[str, Any])
available_tasks (List[Task])
assigned_task_ids (Set[str])

async src.core.ai_powered_task_assignment.find_optimal_task_for_agent_ai_powered(agent_id, agent_status, project_tasks, available_tasks, assigned_task_ids, ai_engine)[source]#

AI-powered task assignment to replace the basic version.

This should be called from request_next_task in marcus_mcp_server.py

Return type:

Optional[Task]

Parameters:

agent_id (str)
agent_status (Dict[str, Any])
project_tasks (List[Task])
available_tasks (List[Task])
assigned_task_ids (Set[str])
ai_engine (MarcusAIEngine)