sadd:launch-sub-agent

What problem does it solve?

Launch an intelligent sub-agent framework that automatically selects optimal model configurations and specialized agents to tackle complex tasks while ensuring isolated context and traceable decision-making.

Core Features & Use Cases

• Automatic model selection based on task complexity (opus, sonnet, haiku) with optional specialized agents.
• Zero-shot Chain-of-Thought reasoning at the start and mandatory self-critique verification at the end.
• Supervisor/Orchestrator pattern for context isolation and modular task execution across sub-agents.
• Supports tasks across development, research, documentation, architecture, and testing with clear runbooks.

Quick Start

Provide a task description to trigger the sub-agent launcher and it will automatically select a model and agent and dispatch a sub-agent.