Which models are used for subtasks?

The skill automatically selects the best model for the job: Opus for complex reasoning, Sonnet for moderate tasks, and Haiku for simple searches and exploration.

Can I run multiple tasks simultaneously?

Yes, the skill is designed for parallel dispatch, allowing you to launch multiple specialized subagents at once to handle different parts of a project concurrently.

How does Subagent Teams improve Claude's performance?

It offloads heavy processing to subagents, keeping the main context window lean and preventing the performance degradation that occurs when AI context windows auto-compact.

When is it best to use this skill?

Use it for complex multi-step tasks, when your context window is getting large, or when you have independent subtasks that don't rely on each other to start.

Does this replace the main Claude decision-making process?

No, the main orchestrator still handles final synthesis and critical decision-making; the subagents act as specialized workers for execution and research.

Subagent Teams Orchestrator

Name: Subagent Teams Orchestrator
Author: alijilani-dev

byalijilani-dev

Productividad y Flujo de Trabajo

Orchestrates complex development tasks by delegating work to parallel specialized subagents to optimize context window performance.

Acerca de

Subagent Teams is a performance-focused skill designed to handle large-scale projects by decomposing complex requests into independent subtasks. By delegating heavy lifting—such as codebase exploration, research, and testing—to specialized subagents running in separate context windows, it prevents the main orchestrator's context from becoming bloated. This isolation prevents 'auto-compact' triggers that degrade AI quality, allowing for consistent, high-precision results while utilizing parallel execution to speed up development workflows.

Características Principales

Intelligent task decomposition and dependency mapping
Unified synthesis of distributed subagent outputs
0 GitHub stars
Context isolation to maintain high-quality reasoning
Parallel subagent dispatch for concurrent execution
Dynamic model selection (Opus, Sonnet, Haiku) based on task complexity

Casos de Uso

Executing massive codebase refactors across multiple directories
Parallelizing feature implementation alongside test suite generation
Conducting deep research and architectural analysis without losing main thread focus

Acerca de

Características Principales

Intelligent task decomposition and dependency mapping
Unified synthesis of distributed subagent outputs
0 GitHub stars
Context isolation to maintain high-quality reasoning
Parallel subagent dispatch for concurrent execution
Dynamic model selection (Opus, Sonnet, Haiku) based on task complexity

Casos de Uso

Executing massive codebase refactors across multiple directories
Parallelizing feature implementation alongside test suite generation
Conducting deep research and architectural analysis without losing main thread focus