v2.1.113
Claude CodeSummary
This release introduces new network sandboxing controls via `sandbox.network.deniedDomains` and enables Remote Control clients to query file autocomplete suggestions. Numerous stability fixes address issues with session archiving, transcript display, markdown pasting, and various error conditions across CLI and UI interactions.
New Features
- Added `sandbox.network.deniedDomains` setting to block specific domains even when a broader `allowedDomains` wildcard would otherwise permit them
- Enabled Remote Control clients to query `@`-file autocomplete suggestions
- Improved `/ultrareview` with faster launch via parallelized checks, inclusion of diffstat in the launch dialog, and an animated launching state
Bug Fixes
- Fixed concurrent-call timeout handling where a message for one tool call could silently disarm another call's watchdog
- Fixed Cmd-backspace / `Ctrl+U` to correctly delete from the cursor to the start of the line
- Fixed markdown tables breaking when a cell contains an inline code span with a pipe character
- Fixed session recap auto-firing while composing unsent text in the prompt
- Fixed `/copy` "Full response" not aligning markdown table columns correctly for pasting into GitHub, Notion, or Slack
- Fixed messages typed while viewing a running subagent being hidden from its transcript and misattributed to the parent AI
- Fixed Bash `dangerouslyDisableSandbox` running commands outside the sandbox without triggering a permission prompt
- Fixed `/effort auto` confirmation message to correctly state "Effort level set to max"
- Fixed the "copied N chars" toast overcounting emoji and other multi-code-unit characters
- Fixed `/insights` crashing with `EBUSY` on Windows
- Fixed exit confirmation dialog mislabeling one-shot scheduled tasks as recurring; it now correctly shows a countdown
- Fixed slash/@ completion menu not sitting flush against the prompt border in fullscreen mode
- Fixed `CLAUDE_CODE_EXTRA_BODY` `output_config.effort` causing 400 errors on subagent calls to models that don't support effort and on Vertex AI
- Fixed prompt cursor disappearing when `NO_COLOR` is set
- Fixed `ToolSearch` ranking so pasted MCP tool names surface the actual tool instead of description-matching siblings
- Fixed compacting a resumed long-context session failing with "Extra usage is required for long context requests"
- Fixed `plugin install` succeeding when a dependency version conflicts with an already-installed plugin; it now correctly reports `range-conflict`
- Fixed "Refine with Ultraplan" not showing the remote session URL in the transcript
- Fixed SDK image content blocks that fail to process from crashing the session; they now degrade to a text placeholder
- Fixed Remote Control sessions not streaming subagent transcripts
- Fixed Remote Control sessions not being archived when Claude Code exits
- Fixed `thinking.type.enabled is not supported` 400 error when using Opus 4.7 via a Bedrock Application Inference Profile ARN
Improvements
- Changed the CLI to spawn a native Claude Code binary (via a per-platform optional dependency) instead of bundled JavaScript
- Enabled viewport scrolling in Fullscreen mode using Shift+↑/↓ when extending a selection past the visible edge
- Enabled `Ctrl+A` and `Ctrl+E` to move to the start/end of the current logical line in multiline input, matching readline behavior
- Enabled `Ctrl+Backspace` to delete the previous word on Windows
- Ensured long URLs in responses and bash output remain clickable when they wrap across lines (in terminals with OSC 8 hyperlinks)
- Improved `/loop` functionality: pressing Esc now cancels pending wakeups, and wakeups display as "Claude resuming /loop wakeup" for clarity
- Enabled `/extra-usage` to work from Remote Control (mobile/web) clients
- Improved handling of subagents that stall mid-stream by failing them with a clear error after 10 minutes instead of hanging silently
- Improved Bash tool transcript display so multi-line commands whose first line is a comment now show the full command, closing a UI-spoofing vector
- Prevented permission prompts when running `cd <current-directory> && git …` where the `cd` command is a no-op
- Enhanced security by treating `/private/{etc,var,tmp,home}` paths as dangerous removal targets under `Bash(rm:*)` allow rules on macOS
- Enhanced security by making Bash deny rules match commands wrapped in `env`, `sudo`, `watch`, `ionice`, `setsid`, and similar exec wrappers
- Enhanced security by preventing `Bash(find:*)` allow rules from auto-approving `find -exec` or `-delete`