Follow Discord
Sweep 30 Oct 2026 · 21:36Z Build v2.1.315 533 read Stable v2.1.308 Latest v2.1.314 Next v2.1.315 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.315 ·

New CLAUDE_CODE_SUBAGENT_EFFORT sets a default effort for subagents

Set CLAUDE_CODE_SUBAGENT_EFFORT to give subagents and workflow agents an effort level when their own definition does not set one

Use it now Notable No documentation found New Features
JSON All of v2.1.315
Use it nowTier: how much it should matter to you
5Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
SubagentsArea: what it touches
New FeaturesKind: in v2.1.315,
What probably matters to youSection of the release

Unclear Which effort levels the variable accepts, other than inherit, is not stated.

What

Subagents are helper agents Claude starts to handle part of a task. Effort is how much reasoning the model puts into its work. Before, a subagent took the session's effort unless its own definition set one.

The new environment variable CLAUDE_CODE_SUBAGENT_EFFORT changes that:

  • When an agent definition has no effort of its own, the level set in the variable is used.
  • The value inherit is accepted.
  • Any other value Claude Code does not recognise is recorded as an unknown level.

The documentation for Workflow scripts now says opts.effort defaults to the configured subagent default when this is set, instead of the session's effort. The variable can also be set through the environment section of Claude Code's settings, like CLAUDE_CODE_SUBAGENT_MODEL_FORCE.

Why

One variable can now limit how much reasoning, and so how much cost, all your subagents use.

Read from
Names in the bundleCLAUDE_CODE_SUBAGENT_EFFORT
How sure we are
Something disagreesSomething we can check disagrees with this entry, or the writer said they could not settle it.
The writer flagged doubtWhich effort levels the variable accepts, other than `inherit`, is not stated.
The name it cites is new in this buildNew in this build: CLAUDE_CODE_SUBAGENT_EFFORT

See this entry in the whole of v2.1.315 →

Feedback