Jack Wotherspoon, Developer Advocate. Software Engineer Abhi Patel. Subagents enable Gemini CLI to assign complex, repetitive, or large-scale tasks to specialized expert agents. Each subagent functions independently, with its own dedicated context window, tailored system prompts, and specialized set of tools. This ensures your primary session stays fast, streamlined, and centered on the overall vision, while delegating the intermediate tasks to a group of specialized subagents. Sorry, your browser doesn’t support video playback. What are subagents?. Subagents are specialized expert agents that work in parallel with your main Gemini CLI session. When given a broad or complex task, Gemini CLI functions as a strategic orchestrator, breaking it down and assigning specific sub-tasks to the most suitable subagents. Each subagent operates independently, equipped with its own tools, MCP servers, system instructions, and context window.
Google Developers Blog