OpenCode Subagents

Overview

OpenCode subagents are specialized AI agents that can be defined and invoked for focused, domain-specific tasks. This skill should be invoked when you need specialized agents with clean context for focused execution, such as code review specialists, testing agents, or documentation agents.

Core Principles

• Single Responsibility: Each subagent should have a clear, focused purpose
• Clean Context: Subagents receive focused context, avoiding pollution from unrelated tasks
• Skill Isolation: Subagents carry their own skills and tools, independent of parent session
• Composition: Multiple subagents can work in parallel on independent tasks

Preparation Checklist

• Define the subagent's purpose and scope of expertise
• Identify the skills the subagent needs to carry
• Determine which tools the subagent should have access to
• Choose appropriate model preferences if different from default