Custom subagents in Claude Code are specialized AI assistants that can be invoked to handle specific types of tasks. They enable more efficient problem-solving by providing task-specific configurations with customized system prompts, tools and a separate context window. When Claude Code encounters a task that matches a subagent’s expertise, it can delegate that task to the specialized subagent, which works independently and returns results.

Built-in agents are general-purpose, statusline-setup, output-style-setup, when type /agents, only these active agents are shown up. From github: https://github.com/wshobson/agents?tab=readme-ov-file, I can clone these subagents to be used. these agents will pick and choose tools like write and read file, also mcp tools: Subagents can access MCP tools from configured MCP servers. When the tools field is omitted, subagents inherit all MCP tools available to the main thread.

To create and manage subagents, you can do it on system and project level

They adhere to a consistent style, and should I need to create one, I will employ the same format.

The quant-analyst subagent’s md file is

reference to this repo.