Follow Discord
Sweep 25 Sep 2026 · 19:33Z Build v2.1.283 504 read Stable v2.1.274 Latest v2.1.283 Next v2.1.283 Feeds RSS JSON llms.txt llms-full.txt Unofficial

Claude Code v2.1.282 ·

Speculative permission check for remote tools gets an environment variable but stays off

A new environment variable, CLAUDE_CODE_REMOTE_TOOLS_SPECULATIVE_CLASSIFIER, is read, but the check it feeds always answers no, so nothing changes

Nothing to try yet Notable In Development
JSON All of v2.1.282
Nothing to try yetTier: how much it should matter to you
3Useful: my rating, 1 to 5
4Signal: worth watching, 1 to 5
Remote ToolsArea: what it touches
In DevelopmentKind: in v2.1.282,
What probably matters to youSection of the release
What

The check for a speculative permission classifier on remote tools (tools run on another machine) is now connected to the code that would use it. It reads a new environment variable, a setting you put in your shell, CLAUDE_CODE_REMOTE_TOOLS_SPECULATIVE_CLASSIFIER, along with a remote switch. Whatever either of them says, the check still always answers no, so setting the variable has no effect in this build.

The remote switch has answered on for this site's own account in earlier readings. That makes no difference here, because the check ignores it.

Why

The name suggests a planned classifier that would judge permission for remote tool use ahead of time. It is wired in but disabled in the code itself.

Read from
Feature flag
tengu_violin_speculative_classifier Off by default, switched on for this account

The shipped code defaults this off, and the flag server returned on for the one account this site reads on this version. That is the reading that makes the entry above worth a second look, and it still says nothing about your account.

This account: on · anonymous baseline: on · compiled default in v2.1.282: off

These values were read against a different version of Claude Code, so treat them as the nearest reading available instead of one taken on this release.

Read once, for one account on one subscription tier, against v2.1.282. It isn't a statement about your account. What a flag value here can and cannot tell you

What has happened since
Flag reading moved The flag server now returns off for tengu_violin_speculative_classifier, read as this account. A reading is one sample. Claude Code evaluates its flags remotely, so no client sees the targeting rule behind a value and this says nothing about your account.

See this across every release →

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 doubtWhat the classifier path would do once enabled is not shown.
The name it cites is new in this buildNew in this build: tengu_violin_speculative_classifier

See this entry in the whole of v2.1.282 →

Feedback