Claude Code v2.1.193: Auto Mode Controls, Telemetry Enhancements, and Background Agent Fixes
Claude Code v2.1.193 adds a shell-wide auto-mode classifier setting, assistant response OpenTelemetry logging, bash path autocomplete, and a batch of fixes for background sessions and agents.
The Claude Code development team has released version 2.1.193, expanding auto-mode configurability, adding new observability capabilities, and resolving a series of issues affecting background sessions and agents. This update gives administrators finer control over shell command classification while improving day-to-day ergonomics like bash path autocomplete and MCP authentication handling.
What’s New
Auto Mode Configuration and Transparency
A new autoMode.classifyAllShell setting routes all Bash and PowerShell commands through the auto-mode classifier, rather than only commands matching arbitrary-code-execution patterns. In addition, auto-mode denial reasons now appear in the transcript, the denial toast, and the recent denials list in /permissions, making it much easier to understand why a command was blocked.
OpenTelemetry Assistant Response Logging
A new claude_code.assistant_response OpenTelemetry log event contains the model’s response text. It is redacted unless OTEL_LOG_ASSISTANT_RESPONSES=1; when that variable is unset it follows OTEL_LOG_USER_PROMPTS, so deployments that already log prompt content will start receiving response content on upgrade. Set OTEL_LOG_ASSISTANT_RESPONSES=0 to keep prompts-only logging.
Terminal and Shell Ergonomics
Bash mode (!) now offers live file path autocomplete. Idle background shell commands are automatically reaped under memory pressure, which can be disabled with CLAUDE_CODE_DISABLE_BG_SHELL_PRESSURE_REAP=1. A startup notice now appears when MCP servers need authentication, pointing you at /mcp.
Background Session and Agent Fixes
This release resolves several backgrounding issues:
- Backgrounding (←←) no longer spuriously cancels with “N background tasks would be abandoned” when all running tasks carry over to the new session
- Pinned background agents are no longer re-prompted to “Continue from where you left off” after every auto-update
- Backgrounding the main turn no longer spawns a phantom “general-purpose (resumed)” subagent that re-ran the main conversation
- The agent panel no longer hides sibling agents when viewing a subagent
- The background agent launch result no longer instructs Claude to “end your response” — it keeps working on other tasks while the agent runs
MCP and Plugin Improvements
MCP headersHelper authentication now re-runs the helper and reconnects automatically when a tool call returns 401/403. Plugin auto-rename now follows marketplace renames maps automatically, updating your settings to the new name.
Other Fixes
/model and other client-data-gated UI no longer show stale or empty state immediately after /login, and the /add-dir message is improved when the directory is already a working directory.
How to Update
To update to Claude Code v2.1.193:
- Check your current version:
claude --version - Update using your preferred method:
- Package manager: Follow your distribution’s update process
- Direct download: Visit the release page for platform-specific binaries
- If you export OpenTelemetry logs, review your
OTEL_LOG_ASSISTANT_RESPONSESsetting before upgrading - Explore
autoMode.classifyAllShellif you want all shell commands routed through the auto-mode classifier