Can My Agent Uselookup · compare · cite

Aug 28, 2026 · Catalog foundation is up

Product feature · runtime

Subagent managementCommon product term

List, inspect, and control child-agent tasks.

Terminology basis: Common product term. Anthropic — Create custom subagents

Subagent management: 0 supported, 5 partial, 0 unsupported, 26 unreviewed across 31 cataloged products.

Markdown · JSON

Explore this familyMore in Subagents9 capabilities

Current evidence by product

Can my agent use Subagent management?

Read across for the answer. 5 of 31 current product columns have reviewed evidence; unreviewed does not mean unsupported.

Known issues (0)Report an issue
  • Supported0
  • Partial5
  • Unsupported0
  • Unknown26
  • Not applicable0

Web

9 products

Desktop

13 products

CLI

9 products

Unknown means no public evidence has been reviewed for that product and capability. It does not mean unsupported.

How statuses are assigned

Definition and scope

What this capability means

This row tracks supervision of child-agent executions after delegation. At minimum, evidence should show how active and queued children are identified and whether cancellation propagates to their tools and descendants.

Record pause, resume, retry, deadline, timeout, priority, log inspection, partial-result recovery, and orphan cleanup separately. A UI that hides a child after cancel while its tools continue running is only partial support.

Traceable compatibility

Assertion ledger

Documentation evidence only. No runtime conformance test is implied.

ChatGPTweb · current
Partial
Target
2026-08-28 ChatGPT Work web documentation observation · hosted-observation
Environment
hosted-default
Observed
2026-08-28
  • runtimeActive and Done lists and completed-result inspection are documented; the web sidebar cannot stop or steer an individual child
Evidence
ChatGPTdesktop · current
Partial
Target
current ChatGPT desktop documentation · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtimechild threads are inspectable and the operator can ask Codex to steer, stop, or close them; pause, retry, deadlines, and priority are not established
Evidence
Codex CLIcli · current
Partial
Target
current Codex CLI documentation · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtime/agent switches among active threads and Codex accepts steer, stop, and close requests; pause, retry, deadlines, and priority are not established
Evidence
Claude CLIcli · current
Partial
Target
current Claude Code subagent and parallel-agent documentation · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtimelist, inspect, background, attach, stop, resume, and maximum-turn controls are documented; a general pause operation and explicit retry contract are not established
Evidence
Copilot Chatdesktop · current
Partial
Target
current VS Code subagent documentation · dated-documentation
Environment
local-default
Observed
2026-08-28
  • runtimelive inspection and persistent read-only child records are documented; follow-up, resume, pause, retry, timeout, and cancellation controls are not established
Evidence
  1. 1. Evidence checked 2026-08-28: Claude Code lists live subagents and background tasks, supports foreground/background execution, stopping and attaching to background work, resuming a subagent by ID, and a per-subagent `maxTurns` limit. Pause and retry semantics are not fully documented.
  2. 2. Evidence checked 2026-08-28: ChatGPT Work exposes read-only Active and Done child lists; local Codex lets operators inspect child threads and ask the parent to steer, stop, or close them, with CLI thread switching through /agent.
  3. 3. Evidence checked 2026-08-28: VS Code exposes each running child, current tool, elapsed time, prompt, full tool-call details, and returned result as a persistent read-only child chat; each child invocation is stateless and cannot receive follow-up messages.