CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API
A Claude Code environment variable, read out of the shipped bundle. It has been in the build since v2.1.321, including the newest one read.
Setting CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API gives remote sessions a separate failure type when their route is refused
Not Anthropic's. This is the line from the earliest changelog entry here that named it, v2.1.321.
In the changelogs
7Releases whose published page names CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API.
-
v2.1.321
New environment variable for user-delegated sessions with its own failure type
Setting CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API gives remote sessions a separate failure type when their route is refused
found in this entry's text, named in this entry
-
v2.1.321
Settings for user delegated sessions are passed on, and some commands refuse to run in them
Two user delegated session variables are passed to spawned processes, and
--bgandplugin evalrefuse to run when the direct API one is setfound in this entry's text
-
v2.1.321
Auto mode explains when no allowed model can check an action
Auto mode now refuses an action with a clear message when none of the models allowed to check it can be used
found in this entry's text, named in this entry
-
v2.1.321
Auto mode's safety checker behaves differently for user-delegated direct API sessions
When
CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_APIis set, auto mode switches off a third-party checker role and always checks allowed MCP write toolsfound in this entry's text, named in this entry
-
v2.1.321
Clear reasons when a session can't run on this machine
Claude Code now explains why a session is refused with a "Your session can't run on this machine" message and a specific cause
found in this entry's text
-
v2.1.321
Unreviewed subagent output now comes with a security warning
Claude Code now warns when a subagent's output was not reviewed, with wording that depends on CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API
found in this entry's text, named in this entry
-
v2.1.321
Auto mode refuses to start a subagent when its safety check is unavailable
In auto mode, a subagent is not started if the safety classifier cannot check its prompt, and Claude Code says why
found in this entry's text
First cited
1The earliest release whose published evidence quoted CLAUDE_CODE_USER_DELEGATED_SESSION_DIRECT_API was v2.1.321, 9 Nov 2026. That is the oldest release this project wrote about it, so it is a floor on the name's age and not the release that introduced it. It is not a presence reading: which builds carried the name is the table below.
Presence across releases
1Build by build
2One row per release since the first build this name was read out of. Absent means the build was read and the name was not in it, never mined means no bundle for that release was ever archived, and a declared type or a default is only ever what that release's own bundle stated.
Read out of the published npm bundle release by release, and out of Anthropic's own documentation as this site captured it. Nothing on this page is a description anybody here wrote about what the environment variable does. All environment variables.