Claude Code v2.1.160
Released Jun 2, 2026 (2 months ago).
171 issues reference this build
12 still open
159 resolved (93%)
"References this build" means the issue body mentions running 2.1.160 (from the
bug-report template). It reflects report volume, not a proven defect count — an issue may mention a
version without being caused by it. First-party data from the anthropics/claude-code tracker.
Top issues reported on v2.1.160
Ranked by community engagement (reactions and comments).
- [BUG] Left arrow key clears entire conversation with no confirmation and no undo
- [BUG] API Error 400 - The request body is not valid JSON: str is not valid UTF-8: surrogates not allowed - mid-conversation error
- [BUG] bwrap sandbox broken on merged-usr systems (Arch): "Can't mount tmpfs on /newroot/lib64" — enableWeakerNestedSandbox does not fix it, MCP servers fail to start
- [FEATURE] Cowork: allow switching model mid-conversation (model is locked after the first message)
- [BUG] Intermittent mid-stream SSE stall — Opus 4.7/4.8, Sonnet 4.6, pause token emission 40–600s mid-response, latency increasing (VS Code extension)
- [Bug] Claude refuses to fill in password fields in testing environment
- Web (claude.ai/code): session groups don't persist across environments/devices — sessions appear "Ungrouped"
- Bash tool falsely reports 'temp filesystem … is full (0MB free)' on any empty-stdout + non-zero-exit command
- [BUG] Desktop App: AskUserQuestion tool hides most of the message
- [BUG] claude-in-chrome MCP can't distinguish two Chrome profiles — `list_connected_browsers` returns unstable generic "Browser 1/2" names, forcing constant disambiguation and wrong-profile risk
- [FEATURE] Cowork: give users a manual /compact (user-initiated compaction)
- Phantom user turn carried a context-aware prompt-injection / data-exfiltration payload (Windows, CLI 2.1.160, Opus 4.8)
- [BUG] macOS: Claude Desktop can't run hooks/git in TCC-protected folders (~/Documents) — `disclaimer` disclaims TCC responsibility, git getcwd() → "Operation not permitted"
- [Bug] GitHub MCP Connection Failed
- [BUG] Desktop + Bedrock: Haiku 4.5 sent as bare model ID (not inference profile) on Scheduled Task follow-ups → intermittent 400 "invalid model identifier"
- [BUG] Slash command does not work with modified keybindings
- WorktreeCreate hook fails with EPERM for repos under ~/Documents — hook spawned via 'disclaimer' shim loses the app's TCC Documents grant
- [Feature Request] Disable automatic branch creation for Claude Code operations
- [BUG] ScheduleWakeup persists after Ctrl+C, daemon auto-respawns loop unattended causing unbounded token spend
- [BUG] FleetView PR chips never clear — keyed by branch name, collide across repos
- Autocompact silently disabled for Bedrock/Vertex users since v2.1.154
- [BUG] Claude in Chrome (VS Code 2.1.160): `mcp_set_servers` publishes only SDK tools to the model and gates even that on `sdkServersChanged` — so a dynamically-added stdio server connects (status/pill shows its tools) but the model never receives them
- Slash-command menu returns zero matches when query contains `.` (regression 2.1.160 → 2.1.161)
- [BUG][Desktop][macOS 26] Code tab kills entire app — rootfs.img redownload loop (disk-write jetsam); crashes even with all connectors disabled
- [Feature Request] Remove unnecessary verbal disclaimers.. All of them. They waste energy, time, tokens and money. Stop the rot.
- [Regression] Thinking blocks no longer expand on click in VSCode extension (v2.1.160)
- [Bug] MCP server not properly initialized - /doctor command fails to detect configuration errors
- VS Code/Cursor: Claude panel restores blank after Remote-SSH reconnect (deserializeWebviewPanel does not rehydrate conversation)
- [Bug] Tool Call Parsing Failed - Unable to Parse Model Response
- [BUG] command autocomplete only triggers at position 0, while @ triggers mid-prompt
Changelog
What's changed
- Added a prompt before writing to shell startup files (
.zshenv,.zlogin,.bash_login) and~/.config/git/, which could otherwise lead to unintended command execution acceptEditsmode now prompts before writing build-tool config files that grant code execution (.npmrc,.yarnrc*,bunfig.toml,.bazelrc,.pre-commit-config.yaml,.devcontainer/, etc.)- Edit no longer requires a separate Read after viewing a file with
grep: single-filegrep/egrep/fgrepcommands now satisfy the read-before-edit check - Fixed copy-on-select not writing to the Windows clipboard on WSL — now uses PowerShell interop instead of OSC 52, which terminals like MobaXterm don't support
- Fixed restoring a completed session from
claude agentsdropping chat history and re-running the original prompt - Fixed background sessions re-attached after overnight retire losing their conversation and re-running the original prompt
- Fixed
claude --bgoccasionally failing with "socket missing" when the background daemon was cold-starting on a loaded machine - Fixed an issue on Windows where the directory a background session was started in could not be deleted after
claude rmuntil the background daemon exited - Fixed background agents that resumed work being shown under Completed in the agents list
- Fixed
claude agentsfreezing for several seconds when returning to the session list due to the auto-updater re-checking on every exit - Fixed Esc, arrow keys, and typing becoming unresponsive on Windows when attached to a background session or in the agent view while the host is under heavy CPU load
- Fixed background agents emitting terminal sync-output markers to terminals that don't support them (Apple Terminal, tmux), causing render artifacts when entering a running agent
- Fixed mouse wheel scrolling prompt history instead of the transcript right after opening a session from the agents list
- Fixed CJK IME composition appearing at the bottom-left of the screen instead of at the input caret in the
claude agentsview - Fixed valid
file:///C:/...links being rewritten to a broken path on Windows terminals with hyperlink support - Fixed voice mode failing to connect when the project directory or branch name contains non-ASCII or special characters
- Fixed the auto mode unavailability message on third-party providers (Bedrock/Vertex/Foundry) to point to the
CLAUDE_CODE_ENABLE_AUTO_MODEopt-in instead of incorrectly blaming the model - Fixed
/effort ultracodeincorrectly blaming the dynamic workflows setting when the model cannot run xhigh; ultracode is no longer offered on models that do not support it - Fixed model-not-found errors suggesting
--modelwhen running via the SDK or other hosts where the CLI flag doesn't apply - Fixed Claude's past replies disappearing from scrollback when resuming a brief mode session with brief mode turned off
- Fixed vim mode
ppasting on the line below instead of at the cursor when the register was yanked withv$ - Improved performance of opening recently-inactive background agent sessions in
claude agents - Improved auto mode classifier latency by reducing reasoning on routine actions, lowering the chance of "could not evaluate this action" blocks
- Improved background-session teardown (
claude rm/stop, idle reap) to send SIGTERM to running shell subprocesses before SIGKILL, so cleanup handlers run - Removed
CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE; the environment variable is now a no-op - Removed the JetBrains plugin install suggestion from startup
- Renamed the dynamic-workflow trigger keyword from
workflowtoultracode. The word "workflow" no longer triggers a run; asking for one in your own words still works. The trigger keyword is highlighted in violet in the prompt input