orch

Use this agent for long-running or multi-session tasks that require maintaining context across conversations. This agent specializes in continuing previous work, resuming interrupted sessions, and managing complex tasks that span multiple interactions. Use proactively when: (1) User references previous work or sessions, (2) Tasks are too large for a single session, (3) User asks to "continue" or "resume" work, (4) Context needs to be preserved across agent handoffs.

Use this agent when you need to break down complex tasks into multiple sub-tasks that can be executed in parallel by specialized agents. Examples: <example> Context: User needs to build a full-stack feature with database schema, API endpoints, and frontend components. user: "I need to create a user authentication system with login, registration, and password reset" assistant: "This is a complex task that requires multiple specialized components. Let me use the parallel-task-orchestrator agent to break this down and coordinate parallel execution." <Task tool call to parallel-task-orchestrator> </example> <example> Context: User is working on a project that requires documentation, testing, and code refactoring simultaneously. user: "Can you help me improve this codebase? I need better docs, tests, and performance optimizations" assistant: "This requires multiple specialized tasks that can be executed in parallel. I'm going to use the parallel-task-orchestrator agent to create a structured plan and coordinate the work." <Task tool call to parallel-task-orchestrator> </example> <example> Context: Proactive intervention when detecting a complex multi-faceted request. user: "I need to migrate our database and update all the API calls" assistant: "I notice this involves multiple independent but related tasks. Let me engage the parallel-task-orchestrator agent to create an efficient parallel execution strategy." <Task tool call to parallel-task-orchestrator> </example>

Use this agent when facing complex, multi-faceted tasks that can benefit from parallel execution through delegation to specialized sub-agents. This includes scenarios requiring coordination of multiple independent workstreams, large-scale refactoring projects, comprehensive codebase analysis, or any task where breaking down work into parallel subtasks would improve efficiency. Examples: <example> Context: User needs to implement a new feature that spans multiple parts of the codebase. user: "Add user authentication with OAuth2 support, including frontend login UI, backend API endpoints, and database schema changes" assistant: "This is a complex feature that can be parallelized effectively. Let me use the task-orchestrator agent to plan and coordinate this work." <commentary> Since this task involves multiple independent components (frontend, backend, database), the task-orchestrator agent should break this into parallel workstreams and delegate to appropriate sub-agents. </commentary> </example> <example> Context: User requests a comprehensive code quality improvement. user: "Review and improve the entire authentication module - check for security issues, add tests, and update documentation" assistant: "I'll use the task-orchestrator agent to plan parallel execution of security review, test creation, and documentation updates." <commentary> The three concerns (security, testing, documentation) are largely independent and can be addressed in parallel by specialized agents. </commentary> </example> <example> Context: User needs multiple files analyzed or modified simultaneously. user: "Update all API endpoints to use the new response format and ensure they all have proper error handling" assistant: "Let me engage the task-orchestrator agent to identify all affected endpoints and coordinate parallel updates." <commentary> Multiple files need similar changes - the orchestrator can identify the scope and delegate parallel modifications. </commentary> </example>

../../../agents/orchestration/context.md

../../../agents/orchestration/parallel.md

../../../agents/orchestration/task.md