---
title: "Subagent delegation"
canonical: "https://canmyagentuse.com/features/subagent-delegation"
contentKind: "feature"
locale: "en"
description: "Delegate a bounded task to a product-native child agent."
llmSummary: "Basic delegation establishes one child-agent boundary but not recursion, MCP access, or write access."
publishedAt: "2026-08-28T00:00:00.000Z"
updatedAt: "2026-08-28T00:00:00.000Z"
verifiedAt: "2026-08-28"
tags: ["runtime","subagents","delegation"]
---

# Subagent delegation

Basic delegation establishes one child-agent boundary but not recursion, MCP access, or write access.

- HTML: https://canmyagentuse.com/features/subagent-delegation
- JSON: https://canmyagentuse.com/api/v1/features/subagent-delegation.json
- Markdown: https://canmyagentuse.com/features/subagent-delegation.md

Terminology basis: **Common product term** — https://github.com/different-ai/openwork/blob/dev/ARCHITECTURE.md.

## Current support at a glance

Subagent delegation: 9 supported, 1 partial, 0 unsupported, 21 unreviewed across 31 cataloged products.

- Reviewed current products: 10 of 31
- Supported: 9
- Partial: 1
- Unsupported: 0
- Unreviewed: 21
- Not applicable: 0

Unknown or unreviewed means insufficient published evidence; it does not mean unsupported.

This row establishes that one agent can create a distinct child execution boundary for a bounded task. A renamed tool call, background shell command, or single-model chain-of-thought step is not enough unless the product exposes a child-agent execution with its own task state.

Deeper recursion, simultaneous children, inherited context, model choice, permissions, supervision, and result transfer are separate sibling capabilities.

## Catalog context

- Category: [runtime](/categories/runtime.md)
- Terminology basis: Common product term
- Aliases: child agent, task delegation
- Family: [Subagents](/features/subagents.md)
- Siblings: [Concurrent subagents](/features/subagent-concurrency.md), [Nested subagents](/features/nested-subagents.md), [Per-subagent model selection](/features/subagent-model-selection.md), [Subagent context](/features/subagent-context-isolation.md), [Subagent management](/features/subagent-lifecycle-controls.md), [Subagent MCP access](/features/subagent-mcp-access.md), [Subagent permissions](/features/subagent-approval-boundaries.md), [Subagent results](/features/subagent-result-handoff.md), [Subagent write access](/features/subagent-write-access.md)

## Compatibility assertions

Unknown means insufficient published evidence; it does not mean unsupported.

### ChatGPT (web)

- Harness: [ChatGPT](/harnesses/chatgpt-web.md)
- current: **Supported**
  - Target: hosted-observation — 2026-08-28 ChatGPT Work web documentation observation; observed 2026-08-28
  - Environment: hosted-default
  - Constraint (plan): subagent workflows and visible activity require an eligible account; Ultra may delegate proactively while other levels generally require an explicit request
  - Evidence: [OpenAI — Subagents in ChatGPT and Codex](https://learn.chatgpt.com/docs/agent-configuration/subagents) — documented; observed 2026-08-28
  - Qualification note 4: Evidence checked 2026-08-28: ChatGPT Work and current local Codex releases can spawn specialized agents, expose their activity, and collect results into the parent response.
- preview: **Unknown**

### Claude (web)

- Harness: [Claude](/harnesses/claude-web.md)
- current: **Unknown**
- preview: **Unknown**

### Gemini (web)

- Harness: [Gemini](/harnesses/gemini-web.md)
- current: **Unknown**
- preview: **Unknown**

### Copilot (web)

- Harness: [Copilot](/harnesses/copilot-web.md)
- current: **Unknown**
- preview: **Unknown**

### Grok (web)

- Harness: [Grok](/harnesses/grok-web.md)
- current: **Unknown**
- preview: **Unknown**

### Grok Bot (desktop)

- Harness: [Grok Bot](/harnesses/grok-bot-desktop.md)
- current: **Partial**
  - Target: hosted-observation — 2026-08-28 Grok Bot desktop documentation observation; observed 2026-08-28
  - Environment: hosted-default
  - Constraint (runtime): delegation uses distinct persistent named Bots with asynchronous messages and visible handoffs, rather than documented ephemeral child sessions
  - Constraint (policy): all Bots for an account share one hosted computer, so delegation does not create a separate filesystem, login, or credential security boundary
  - Evidence: [xAI — Grok Bot messaging and collaboration](https://docs.x.ai/grok-bot/chat-and-collaboration) — documented; observed 2026-08-28
  - Qualification note 2: Evidence checked 2026-08-28: Grok Bot can create focused named Bots, let Bots message each other asynchronously, and pass work among them. This is persistent peer-Bot delegation, not documented ephemeral child-session spawning.

### Perplexity (web)

- Harness: [Perplexity](/harnesses/perplexity-web.md)
- current: **Unknown**
- preview: **Unknown**

### Le Chat (web)

- Harness: [Le Chat](/harnesses/le-chat.md)
- current: **Unknown**
- preview: **Unknown**

### Devin (web)

- Harness: [Devin](/harnesses/devin-web.md)
- current: **Unknown**
- preview: **Unknown**

### Replit Agent (web)

- Harness: [Replit Agent](/harnesses/replit-agent.md)
- current: **Unknown**
- preview: **Unknown**

### ChatGPT (desktop)

- Harness: [ChatGPT](/harnesses/chatgpt-desktop.md)
- current: **Supported**
  - Target: dated-documentation — current ChatGPT desktop documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): current local Codex releases delegate after a direct request or applicable AGENTS.md or skill instruction and surface each child thread
  - Evidence: [OpenAI — Subagents in ChatGPT and Codex](https://learn.chatgpt.com/docs/agent-configuration/subagents) — documented; observed 2026-08-28
  - Qualification note 4: Evidence checked 2026-08-28: ChatGPT Work and current local Codex releases can spawn specialized agents, expose their activity, and collect results into the parent response.
- preview: **Unknown**

### Claude (desktop)

- Harness: [Claude](/harnesses/claude-desktop.md)
- current: **Unknown**
- preview: **Unknown**

### Cursor (desktop)

- Harness: [Cursor](/harnesses/cursor.md)
- current: **Supported**
  - Target: dated-documentation — current Cursor subagent documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): applies to Cursor editor, CLI, and Cloud Agents; named subagents can be invoked with /name or natural language
  - Evidence: [Cursor — Subagents](https://cursor.com/docs/subagents) — documented; observed 2026-08-28
  - Qualification note 6: Evidence checked 2026-08-28: Cursor Agent can automatically or explicitly launch built-in and custom subagents on editor, CLI, and Cloud Agent surfaces.
- preview: **Unknown**

### OpenWork Desktop (desktop)

- Harness: [OpenWork Desktop](/harnesses/openwork-desktop.md)
- current: **Supported**
  - Target: dated-documentation — current OpenWork architecture documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): supplied by the embedded OpenCode agent runtime
  - Evidence: [OpenWork — Architecture](https://github.com/different-ai/openwork/blob/dev/ARCHITECTURE.md) — documented; observed 2026-08-28
  - Qualification note 1: Evidence checked 2026-08-28: OpenWork identifies agents as OpenCode primitives for specialized tasks that may use different models and extra context.

### Copilot Chat (desktop)

- Harness: [Copilot Chat](/harnesses/vscode-copilot.md)
- current: **Supported**
  - Target: dated-documentation — current VS Code subagent documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (host-role): claim is scoped to GitHub Copilot's local VS Code agent surface with the subagent tool enabled
  - Evidence: [Visual Studio Code — Subagents in Visual Studio Code](https://code.visualstudio.com/docs/agents/run/subagents) — documented; observed 2026-08-28
  - Qualification note 8: Evidence checked 2026-08-28: VS Code documents the `agent/runSubagent` tool, built-in and custom child agents, autonomous child work, and final-result return to the main agent.
- preview: **Unknown**

### Chrome WebMCP origin trial (desktop)

- Harness: [Chrome WebMCP origin trial](/harnesses/chrome-webmcp-preview.md)
- current: **Unknown**

### Windsurf (desktop)

- Harness: [Windsurf](/harnesses/windsurf.md)
- current: **Unknown**
- preview: **Unknown**

### Zed Agent (desktop)

- Harness: [Zed Agent](/harnesses/zed-agent.md)
- current: **Unknown**
- preview: **Unknown**

### Continue (desktop)

- Harness: [Continue](/harnesses/continue.md)
- current: **Unknown**
- preview: **Unknown**

### Cline (desktop)

- Harness: [Cline](/harnesses/cline.md)
- current: **Unknown**
- preview: **Unknown**

### JetBrains AI (desktop)

- Harness: [JetBrains AI](/harnesses/jetbrains-ai.md)
- current: **Unknown**
- preview: **Unknown**

### Warp (desktop)

- Harness: [Warp](/harnesses/warp.md)
- current: **Unknown**
- preview: **Unknown**

### Claude CLI (cli)

- Harness: [Claude CLI](/harnesses/claude-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Claude Code subagent documentation; observed 2026-08-28
  - Environment: local-default
  - Evidence: [Anthropic — Create custom subagents](https://code.claude.com/docs/en/sub-agents) — documented; observed 2026-08-28
  - Qualification note 3: Evidence checked 2026-08-28: Claude Code documents built-in and custom subagents that receive delegated tasks through the Agent tool, run in their own context, and return a summary to the parent conversation.
- preview: **Unknown**

### ChatGPT CLI (cli)

- Harness: [ChatGPT CLI](/harnesses/chatgpt-cli.md)
- current: **Unknown**
- preview: **Unknown**

### Codex CLI (cli)

- Harness: [Codex CLI](/harnesses/codex-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Codex CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): delegates after a direct request or applicable project or skill instruction; /agent opens child threads while the parent collects their results
  - Evidence: [OpenAI — Subagents in ChatGPT and Codex](https://learn.chatgpt.com/docs/agent-configuration/subagents) — documented; observed 2026-08-28
  - Qualification note 4: Evidence checked 2026-08-28: ChatGPT Work and current local Codex releases can spawn specialized agents, expose their activity, and collect results into the parent response.
- preview: **Unknown**

### OpenCode (cli)

- Harness: [OpenCode](/harnesses/opencode.md)
- current: **Unknown**
- preview: **Unknown**

### Gemini CLI (cli)

- Harness: [Gemini CLI](/harnesses/gemini-cli.md)
- current: **Supported**
  - Target: dated-documentation — current Gemini CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): the main agent can automatically choose a matching subagent or be nudged to a named specialist with @agent syntax
  - Evidence: [Google — Gemini CLI subagents](https://geminicli.com/docs/core/subagents/) — documented; observed 2026-08-28
  - Qualification note 5: Evidence checked 2026-08-28: Gemini CLI exposes built-in and custom subagents as tools, with automatic delegation or explicit @agent prompting.
- preview: **Unknown**

### Aider (cli)

- Harness: [Aider](/harnesses/aider.md)
- current: **Unknown**
- preview: **Unknown**

### Goose (cli)

- Harness: [Goose](/harnesses/goose.md)
- current: **Unknown**
- preview: **Unknown**

### Copilot CLI (cli)

- Harness: [Copilot CLI](/harnesses/copilot-cli.md)
- current: **Supported**
  - Target: dated-documentation — current GitHub Copilot CLI documentation; observed 2026-08-28
  - Environment: local-default
  - Constraint (runtime): the main Copilot agent can automatically invoke built-in or custom agents as temporary subagents
  - Evidence: [GitHub — Custom agents and subagents in Copilot CLI](https://docs.github.com/en/copilot/concepts/agents/copilot-cli/about-custom-agents) — documented; observed 2026-08-28
  - Qualification note 7: Evidence checked 2026-08-28: GitHub Copilot CLI can automatically run built-in or custom agents as temporary subagents with separate context windows.
- preview: **Unknown**

### Amp (cli)

- Harness: [Amp](/harnesses/amp-cli.md)
- current: **Unknown**
- preview: **Unknown**