Skip to content

Claude Code turns on fork subagents by default everywhere

Version 2.1.302 makes forked subagents and background execution the default in claude -p, the Agent SDK, the desktop app and VS Code, not just the terminal.

Source: Claude Code changelog

Claude Code 2.1.302 turns on fork subagents (subagent_type: "fork") by default in claude -p, the Agent SDK, the desktop app and VS Code. The terminal already defaulted to this; now every surface matches it. claude mcp serve never offers forked subagents.

The same release changes how subagents run. The desktop app and VS Code now always run a main conversation’s subagents in the background, the way the terminal does, instead of letting Claude choose. But a subagent’s own nested Agent calls now run in the foreground by default across all four surfaces. To go back to the old forking behaviour:

CLAUDE_CODE_FORK_SUBAGENT=0

Why it matters: if you build on the Agent SDK or drive Claude Code from scripts with claude -p, this changes how your subagents execute without you touching your own code. Forked subagents and background execution were terminal-only quirks; now they’re the baseline everywhere, so behaviour you only tested in the terminal should now hold in the desktop app, VS Code and your own SDK integrations too.

Share

Related reading

ML for Web Devs

Tokens, context windows, and why your bill is what it is

The unit of billing for language models is not the request or the word. Once you understand tokens and how a context window fills up, the invoice stops being a surprise.

ML for Web Devs

RAG is just search with extra steps

Retrieval-augmented generation gets written up like an architecture. It is a search query, a string concatenation and one API call, and the search query is the part that decides whether it works.

ML for Web Devs

A prompt is a spec, so write it like one

Prompt engineering has very little to do with magic words. It is the same skill as writing a clear ticket for a capable contractor who will not ask you any questions.

More from The Wire

All briefs

GitHub Copilot retires six AI models today

Gemini 3.7 Flash, GPT-5.5, GPT-5.4, GPT-5.4 mini, GPT-5 mini and Grok 4.5 stop working across Copilot Chat, inline edits, agent mode and completions from today.

via GitHub