You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
SubagentsArea: what it touches
ImprovementsKind: in v2.1.315,
ImprovementsSection of the release
Unclear Which effort level the permission settings produce is not known.
What
A subagent is a helper Claude starts to handle part of a task. Its effort level sets how much reasoning the model puts in. When a subagent has no valid effort level set, Claude Code now works one out from the layers of permission settings that apply, instead of leaving it unset.
Why
Subagents that used to have no effort level now run at one, which can change how much reasoning they do.
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 level the permission settings produce is not known.
