Orchestrate specialized AI subagents to delegate complex tasks and preserve context within Claude Code workflows.
The Subagent System allows you to create and manage domain-specific AI personalities that Claude Code can delegate work to. Each subagent operates within its own context window, preventing the main conversation from becoming cluttered while applying specialized expertise to specific tasks like testing, code review, or documentation. By configuring custom system prompts and tool permissions for each agent, you can build a highly efficient multi-agent architecture tailored to your project's unique requirements.
주요 기능
01Context preservation through isolated subagent workspaces
02Project-level and global user-level agent management
03Configurable tool permissions and model selection per agent
04Domain-specific expertise with custom system prompts
05Automatic task delegation based on agent descriptions
06491 GitHub stars
사용 사례
01Enforcing architectural standards via a specialized code-reviewer agent
02Managing complex documentation updates without polluting the primary coding context
03Automating test-and-fix loops with a dedicated test-runner subagent
