# Why Use Subagents — Advanced Agent Workflows and Skills

Source: https://www.geekswithgeeks.com/en/agent-workflows/sub-why

> Keep the main context clean by sending noisy research or search work to a separate agent.

## Context isolation

Searching a large codebase fills the context with file contents you only needed once. A **subagent** does that exploring in its own context window and reports back a short answer. The main conversation stays small, focused and cheaper.

## Delegate, isolate, summarise

A subagent works in its own context and returns only a short result to the main agent.

![Three steps: delegate, work in isolation, return a summary.](assets/figures/agent-workflows/section-3-map.svg) — Figure 3.1 — Delegate, isolate and summarise.

## Sending an intern to the library

Instead of reading ten books yourself, you send an intern who returns with a one-page summary. You keep your desk clear for the real decision.

## Do not delegate everything

Each subagent starts with no memory of the conversation, so it needs a self-contained brief. For small tasks the overhead of writing that brief is larger than doing the work directly.

**Quiz:** What is the main benefit of a subagent?

- [x] It keeps noisy work out of the main context
- [ ] It removes the need for tools
- [ ] It makes the model smarter
- [ ] It guarantees correct answers

*Answer:* It keeps noisy work out of the main context. Isolation protects the main context from large intermediate output.