SDK: Subagent System Prompt Injection / Claude Code v2.1.111

What

SDK users can now append custom system prompt content to all Task-tool subagents via a new appendSubagentSystemPrompt option.

Usage

// In SDK session options
{
  appendSubagentSystemPrompt: "Your custom instructions here"
}

Details

The appendSubagentSystemPrompt field is propagated to nested subagents automatically
Gated by the CLAUDE_CODE_ENABLE_APPEND_SUBAGENT_PROMPT environment variable
Allows SDK consumers to inject additional instructions into every subagent without modifying the main system prompt

Evidence

SDK subagent prompt support (search for "appendSubagentSystemPrompt") — 0 occurrences in v2.1.110, 13 in v2.1.111

Strings lifted out of the shipped bundle, so the claim above can be checked against them.

The build it came inv2.1.111

16 Apr 2026 · 23 entries, this one is #6.

Either side of it

Open the release →

Names in this entry

settingappendSubagentSystemPrompt
A new appendSubagentSystemPrompt option lets a subagent's system-prompt append differ from the main session's.
envCLAUDE_CODE_ENABLE_APPEND_SUBAGENT_PROMPT
Gates the extra subagent system prompt: when this is truthy and a prompt was supplied by the hidden --append-subagent-system-prompt option, its file variant, or the SDK's appendSubagentSystemPrompt initialize field, the text is appended to a Task-tool subagent's system prompt (skipped for agents with an isolated context and in one other case). Claude Code sets it to 1 itself when either command-line option supplies a prompt.

Is this right?

Other releases about the same thing. Found by shared names or similar wording; neither means one caused the other.