spacehub

surshar / claude-code

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

Built by Anthropic · github.com/anthropics

claude-code / CHANGELOG.md
GitHub Actionschore: Update CHANGELOG.md b18f2e7 · on Feb 19, 2026 History
1,829 lines · 104.0 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.49
4
5- Fixed Ctrl+C and ESC being silently ignored when background agents are running and the main thread is idle. Pressing twice within 3 seconds now kills all background agents.
6- Fixed prompt suggestion cache regression that reduced cache hit rates.
7- Fixed `plugin enable` and `plugin disable` to auto-detect the correct scope when `--scope` is not specified, instead of always defaulting to user scope
8- Simple mode (`CLAUDE_CODE_SIMPLE`) now includes the file edit tool in addition to the Bash tool, allowing direct file editing in simple mode.
9- Permission suggestions are now populated when safety checks trigger an ask response, enabling SDK consumers to display permission options
10- Sonnet 4.5 with 1M context is being removed from the Max plan in favor of our frontier Sonnet 4.6 model, which now has 1M context. Please switch in /model.
11- Fixed verbose mode not updating thinking block display when toggled via `/config` — memo comparators now correctly detect verbose changes
12- Fixed unbounded WASM memory growth during long sessions by periodically resetting the tree-sitter parser
13- Fixed potential rendering issues caused by stale yoga layout references
14- Improved performance in non-interactive mode (`-p`) by skipping unnecessary API calls during startup
15- Improved performance by caching authentication failures for HTTP and SSE MCP servers, avoiding repeated connection attempts to servers requiring auth
16- Fixed unbounded memory growth during long-running sessions caused by Yoga WASM linear memory never shrinking
17- SDK model info now includes `supportsEffort`, `supportedEffortLevels`, and `supportsAdaptiveThinking` fields so consumers can discover model capabilities.
18- Added `ConfigChange` hook event that fires when configuration files change during a session, enabling enterprise security auditing and optional blocking of settings changes.
19- Improved startup performance by caching MCP auth failures to avoid redundant connection attempts
20- Improved startup performance by reducing HTTP calls for analytics token counting
21- Improved startup performance by batching MCP tool token counting into a single API call
22- Fixed `disableAllHooks` setting to respect managed settings hierarchy — non-managed settings can no longer disable managed hooks set by policy (#26637)
23- Fixed `--resume` session picker showing raw XML tags for sessions that start with commands like `/clear`. Now correctly falls through to the session ID fallback.
24- Improved permission prompts for path safety and working directory blocks to show the reason for the restriction instead of a bare prompt with no context
25
26## 2.1.47
27
28- Fixed FileWriteTool line counting to preserve intentional trailing blank lines instead of stripping them with `trimEnd()`.
29- Fixed Windows terminal rendering bugs caused by `os.EOL` (`\r\n`) in display code — line counts now show correct values instead of always showing 1 on Windows.
30- Improved VS Code plan preview: auto-updates as Claude iterates, enables commenting only when the plan is ready for review, and keeps the preview open when rejecting so Claude can revise.
31- Fixed a bug where bold and colored text in markdown output could shift to the wrong characters on Windows due to `\r\n` line endings.
32- Fixed compaction failing when conversation contains many PDF documents by stripping document blocks alongside images before sending to the compaction API (anthropics/claude-code#26188)
33- Improved memory usage in long-running sessions by releasing API stream buffers, agent context, and skill state after use
34- Improved startup performance by deferring SessionStart hook execution, reducing time-to-interactive by ~500ms.
35- Fixed an issue where bash tool output was silently discarded on Windows when using MSYS2 or Cygwin shells.
36- Improved performance of `@` file mentions - file suggestions now appear faster by pre-warming the index on startup and using session-based caching with background refresh.
37- Improved memory usage by trimming agent task message history after tasks complete
38- Improved memory usage during long agent sessions by eliminating O(n²) message accumulation in progress updates
39- Fixed the bash permission classifier to validate that returned match descriptions correspond to actual input rules, preventing hallucinated descriptions from incorrectly granting permissions
40- Fixed user-defined agents only loading one file on NFS/FUSE filesystems that report zero inodes (anthropics/claude-code#26044)
41- Fixed plugin agent skills silently failing to load when referenced by bare name instead of fully-qualified plugin name (anthropics/claude-code#25834)
42- Search patterns in collapsed tool results are now displayed in quotes for clarity
43- Windows: Fixed CWD tracking temp files never being cleaned up, causing them to accumulate indefinitely (anthropics/claude-code#17600)
44- Use `ctrl+f` to kill all background agents instead of double-pressing ESC. Background agents now continue running when you press ESC to cancel the main thread, giving you more control over agent lifecycle.
45- Fixed API 400 errors ("thinking blocks cannot be modified") that occurred in sessions with concurrent agents, caused by interleaved streaming content blocks preventing proper message merging.
46- Simplified teammate navigation to use only Shift+Down (with wrapping) instead of both Shift+Up and Shift+Down.
47- Fixed an issue where a single file write/edit error would abort all other parallel file write/edit operations. Independent file mutations now complete even when a sibling fails.
48- Added `last_assistant_message` field to Stop and SubagentStop hook inputs, providing the final assistant response text so hooks can access it without parsing transcript files.
49- Fixed custom session titles set via `/rename` being lost after resuming a conversation (anthropics/claude-code#23610)
50- Fixed collapsed read/search hint text overflowing on narrow terminals by truncating from the start.
51- Fixed an issue where bash commands with backslash-newline continuation lines (e.g., long commands split across multiple lines with `\`) would produce spurious empty arguments, potentially breaking command execution.
52- Fixed built-in slash commands (`/help`, `/model`, `/compact`, etc.) being hidden from the autocomplete dropdown when many user skills are installed (anthropics/claude-code#22020)
53- Fixed MCP servers not appearing in the MCP Management Dialog after deferred loading
54- Fixed session name persisting in status bar after `/clear` command (anthropics/claude-code#26082)
55- Fixed crash when a skill's `name` or `description` in SKILL.md frontmatter is a bare number (e.g., `name: 3000`) — the value is now properly coerced to a string (anthropics/claude-code#25837)
56- Fixed /resume silently dropping sessions when the first message exceeds 16KB or uses array-format content (anthropics/claude-code#25721)
57- Added `chat:newline` keybinding action for configurable multi-line input (anthropics/claude-code#26075)
58- Added `added_dirs` to the statusline JSON `workspace` section, exposing directories added via `/add-dir` to external scripts (anthropics/claude-code#26096)
59- Fixed `claude doctor` misclassifying mise and asdf-managed installations as native installs (anthropics/claude-code#26033)
60- Fixed zsh heredoc failing with "read-only file system" error in sandboxed commands (anthropics/claude-code#25990)
61- Fixed agent progress indicator showing inflated tool use count (anthropics/claude-code#26023)
62- Fixed image pasting not working on WSL2 systems where Windows copies images as BMP format (anthropics/claude-code#25935)
63- Fixed background agent results returning raw transcript data instead of the agent's final answer (anthropics/claude-code#26012)
64- Fixed Warp terminal incorrectly prompting for Shift+Enter setup when it supports it natively (anthropics/claude-code#25957)
65- Fixed CJK wide characters causing misaligned timestamps and layout elements in the TUI (anthropics/claude-code#26084)
66- Fixed custom agent `model` field in `.claude/agents/*.md` being ignored when spawning team teammates (anthropics/claude-code#26064)
67- Fixed plan mode being lost after context compaction, causing the model to switch from planning to implementation mode (anthropics/claude-code#26061)
68- Fixed `alwaysThinkingEnabled: true` in settings.json not enabling thinking mode on Bedrock and Vertex providers (anthropics/claude-code#26074)
69- Fixed `tool_decision` OTel telemetry event not being emitted in headless/SDK mode (anthropics/claude-code#26059)
70- Fixed session name being lost after context compaction — renamed sessions now preserve their custom title through compaction (anthropics/claude-code#26121)
71- Increased initial session count in resume picker from 10 to 50 for faster session discovery (anthropics/claude-code#26123)
72- Windows: fixed worktree session matching when drive letter casing differs (anthropics/claude-code#26123)
73- Fixed `/resume <session-id>` failing to find sessions whose first message exceeds 16KB (anthropics/claude-code#25920)
74- Fixed "Always allow" on multiline bash commands creating invalid permission patterns that corrupt settings (anthropics/claude-code#25909)
75- Fixed React crash (error #31) when a skill's `argument-hint` in SKILL.md frontmatter uses YAML sequence syntax (e.g., `[topic: foo | bar]`) — the value is now properly coerced to a string (anthropics/claude-code#25826)
76- Fixed crash when using `/fork` on sessions that used web search — null entries in search results from transcript deserialization are now handled gracefully (anthropics/claude-code#25811)
77- Fixed read-only git commands triggering FSEvents file watcher loops on macOS by adding --no-optional-locks flag (anthropics/claude-code#25750)
78- Fixed custom agents and skills not being discovered when running from a git worktree — project-level `.claude/agents/` and `.claude/skills/` from the main repository are now included (anthropics/claude-code#25816)
79- Fixed non-interactive subcommands like `claude doctor` and `claude plugin validate` being blocked inside nested Claude sessions (anthropics/claude-code#25803)
80- Windows: Fixed the same CLAUDE.md file being loaded twice when drive letter casing differs between paths (anthropics/claude-code#25756)
81- Fixed inline code spans in markdown being incorrectly parsed as bash commands (anthropics/claude-code#25792)
82- Fixed teammate spinners not respecting custom spinnerVerbs from settings (anthropics/claude-code#25748)
83- Fixed shell commands permanently failing after a command deletes its own working directory (anthropics/claude-code#26136)
84- Fixed hooks (PreToolUse, PostToolUse) silently failing to execute on Windows by using Git Bash instead of cmd.exe (anthropics/claude-code#25981)
85- Fixed LSP `findReferences` and other location-based operations returning results from gitignored files (e.g., `node_modules/`, `venv/`) (anthropics/claude-code#26051)
86- Moved config backup files from home directory root to `~/.claude/backups/` to reduce home directory clutter (anthropics/claude-code#26130)
87- Fixed sessions with large first prompts (>16KB) disappearing from the /resume list (anthropics/claude-code#26140)
88- Fixed shell functions with double-underscore prefixes (e.g., `__git_ps1`) not being preserved across shell sessions (anthropics/claude-code#25824)
89- Fixed spinner showing "0 tokens" counter before any tokens have been received (anthropics/claude-code#26105)
90- VSCode: Fixed conversation messages appearing dimmed while the AskUserQuestion dialog is open (anthropics/claude-code#26078)
91- Fixed background tasks failing in git worktrees due to remote URL resolution reading from worktree-specific gitdir instead of the main repository config (anthropics/claude-code#26065)
92- Fixed Right Alt key leaving visible `[25~` escape sequence residue in the input field on Windows/Git Bash terminals (anthropics/claude-code#25943)
93- The `/rename` command now updates the terminal tab title by default (anthropics/claude-code#25789)
94- Fixed Edit tool silently corrupting Unicode curly quotes (\u201c\u201d \u2018\u2019) by replacing them with straight quotes when making edits (anthropics/claude-code#26141)
95- Fixed OSC 8 hyperlinks only being clickable on the first line when link text wraps across multiple terminal lines.
96
97## 2.1.46
98
99- Fixed orphaned CC processes after terminal disconnect on macOS
100- Added support for using claude.ai MCP connectors in Claude Code
101
102## 2.1.45
103
104- Added support for Claude Sonnet 4.6
105- Added support for reading `enabledPlugins` and `extraKnownMarketplaces` from `--add-dir` directories
106- Added `spinnerTipsOverride` setting to customize spinner tips — configure `tips` with an array of custom tip strings, and optionally set `excludeDefault: true` to show only your custom tips instead of the built-in ones
107- Added `SDKRateLimitInfo` and `SDKRateLimitEvent` types to the SDK, enabling consumers to receive rate limit status updates including utilization, reset times, and overage information
108- Fixed Agent Teams teammates failing on Bedrock, Vertex, and Foundry by propagating API provider environment variables to tmux-spawned processes (anthropics/claude-code#23561)
109- Fixed sandbox "operation not permitted" errors when writing temporary files on macOS by using the correct per-user temp directory (anthropics/claude-code#21654)
110- Fixed Task tool (backgrounded agents) crashing with a `ReferenceError` on completion (anthropics/claude-code#22087)
111- Fixed autocomplete suggestions not being accepted on Enter when images are pasted in the input
112- Fixed skills invoked by subagents incorrectly appearing in main session context after compaction
113- Fixed excessive `.claude.json.backup` files accumulating on every startup
114- Fixed plugin-provided commands, agents, and hooks not being available immediately after installation without requiring a restart
115- Improved startup performance by removing eager loading of session history for stats caching
116- Improved memory usage for shell commands that produce large output — RSS no longer grows unboundedly with command output size
117- Improved collapsed read/search groups to show the current file or search pattern being processed beneath the summary line while active
118- [VSCode] Improved permission destination choice (project/user/session) to persist across sessions
119
120## 2.1.44
121
122- Fixed ENAMETOOLONG errors for deeply-nested directory paths
123- Fixed auth refresh errors
124
125## 2.1.43
126
127- Fixed AWS auth refresh hanging indefinitely by adding a 3-minute timeout
128- Fixed spurious warnings for non-agent markdown files in `.claude/agents/` directory
129- Fixed structured-outputs beta header being sent unconditionally on Vertex/Bedrock
130
131## 2.1.42
132
133- Improved startup performance by deferring Zod schema construction
134- Improved prompt cache hit rates by moving date out of system prompt
135- Added one-time Opus 4.6 effort callout for eligible users
136- Fixed /resume showing interrupt messages as session titles
137- Fixed image dimension limit errors to suggest /compact
138
139## 2.1.41
140
141- Added guard against launching Claude Code inside another Claude Code session
142- Fixed Agent Teams using wrong model identifier for Bedrock, Vertex, and Foundry customers
143- Fixed a crash when MCP tools return image content during streaming
144- Fixed /resume session previews showing raw XML tags instead of readable command names
145- Improved model error messages for Bedrock/Vertex/Foundry users with fallback suggestions
146- Fixed plugin browse showing misleading "Space to Toggle" hint for already-installed plugins
147- Fixed hook blocking errors (exit code 2) not showing stderr to the user
148- Added `speed` attribute to OTel events and trace spans for fast mode visibility
149- Added `claude auth login`, `claude auth status`, and `claude auth logout` CLI subcommands
150- Added Windows ARM64 (win32-arm64) native binary support
151- Improved `/rename` to auto-generate session name from conversation context when called without arguments
152- Improved narrow terminal layout for prompt footer
153- Fixed file resolution failing for @-mentions with anchor fragments (e.g., `@README.md#installation`)
154- Fixed FileReadTool blocking the process on FIFOs, `/dev/stdin`, and large files
155- Fixed background task notifications not being delivered in streaming Agent SDK mode
156- Fixed cursor jumping to end on each keystroke in classifier rule input
157- Fixed markdown link display text being dropped for raw URL
158- Fixed auto-compact failure error notifications being shown to users
159- Fixed permission wait time being included in subagent elapsed time display
160- Fixed proactive ticks firing while in plan mode
161- Fixed clear stale permission rules when settings change on disk
162- Fixed hook blocking errors showing stderr content in UI
163
164## 2.1.39
165
166- Improved terminal rendering performance
167- Fixed fatal errors being swallowed instead of displayed
168- Fixed process hanging after session close
169- Fixed character loss at terminal screen boundary
170- Fixed blank lines in verbose transcript view
171
172## 2.1.38
173
174- Fixed VS Code terminal scroll-to-top regression introduced in 2.1.37
175- Fixed Tab key queueing slash commands instead of autocompleting
176- Fixed bash permission matching for commands using environment variable wrappers
177- Fixed text between tool uses disappearing when not using streaming
178- Fixed duplicate sessions when resuming in VS Code extension
179- Improved heredoc delimiter parsing to prevent command smuggling
180- Blocked writes to `.claude/skills` directory in sandbox mode
181
182## 2.1.37
183
184- Fixed an issue where /fast was not immediately available after enabling /extra-usage
185
186## 2.1.36
187
188- Fast mode is now available for Opus 4.6. Learn more at https://code.claude.com/docs/en/fast-mode
189
190## 2.1.34
191
192- Fixed a crash when agent teams setting changed between renders
193- Fixed a bug where commands excluded from sandboxing (via `sandbox.excludedCommands` or `dangerouslyDisableSandbox`) could bypass the Bash ask permission rule when `autoAllowBashIfSandboxed` was enabled
194
195## 2.1.33
196
197- Fixed agent teammate sessions in tmux to send and receive messages
198- Fixed warnings about agent teams not being available on your current plan
199- Added `TeammateIdle` and `TaskCompleted` hook events for multi-agent workflows
200- Added support for restricting which sub-agents can be spawned via `Task(agent_type)` syntax in agent "tools" frontmatter
201- Added `memory` frontmatter field support for agents, enabling persistent memory with `user`, `project`, or `local` scope
202- Added plugin name to skill descriptions and `/skills` menu for better discoverability
203- Fixed an issue where submitting a new message while the model was in extended thinking would interrupt the thinking phase
204- Fixed an API error that could occur when aborting mid-stream, where whitespace text combined with a thinking block would bypass normalization and produce an invalid request
205- Fixed API proxy compatibility issue where 404 errors on streaming endpoints no longer triggered non-streaming fallback
206- Fixed an issue where proxy settings configured via `settings.json` environment variables were not applied to WebFetch and other HTTP requests on the Node.js build
207- Fixed `/resume` session picker showing raw XML markup instead of clean titles for sessions started with slash commands
208- Improved error messages for API connection failures — now shows specific cause (e.g., ECONNREFUSED, SSL errors) instead of generic "Connection error"
209- Errors from invalid managed settings are now surfaced
210- VSCode: Added support for remote sessions, allowing OAuth users to browse and resume sessions from claude.ai
211- VSCode: Added git branch and message count to the session picker, with support for searching by branch name
212- VSCode: Fixed scroll-to-bottom under-scrolling on initial session load and session switch
213
214## 2.1.32
215
216- Claude Opus 4.6 is now available!
217- Added research preview agent teams feature for multi-agent collaboration (token-intensive feature, requires setting CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1)
218- Claude now automatically records and recalls memories as it works
219- Added "Summarize from here" to the message selector, allowing partial conversation summarization.
220- Skills defined in `.claude/skills/` within additional directories (`--add-dir`) are now loaded automatically.
221- Fixed `@` file completion showing incorrect relative paths when running from a subdirectory
222- Updated --resume to re-use --agent value specified in previous conversation by default.
223- Fixed: Bash tool no longer throws "Bad substitution" errors when heredocs contain JavaScript template literals like `${index + 1}`, which previously interrupted tool execution
224- Skill character budget now scales with context window (2% of context), so users with larger context windows can see more skill descriptions without truncation
225- Fixed Thai/Lao spacing vowels (สระ า, ำ) not rendering correctly in the input field
226- VSCode: Fixed slash commands incorrectly being executed when pressing Enter with preceding text in the input field
227- VSCode: Added spinner when loading past conversations list
228
229## 2.1.31
230
231- Added session resume hint on exit, showing how to continue your conversation later
232- Added support for full-width (zenkaku) space input from Japanese IME in checkbox selection
233- Fixed PDF too large errors permanently locking up sessions, requiring users to start a new conversation
234- Fixed bash commands incorrectly reporting failure with "Read-only file system" errors when sandbox mode was enabled
235- Fixed a crash that made sessions unusable after entering plan mode when project config in `~/.claude.json` was missing default fields
236- Fixed `temperatureOverride` being silently ignored in the streaming API path, causing all streaming requests to use the default temperature (1) regardless of the configured override
237- Fixed LSP shutdown/exit compatibility with strict language servers that reject null params
238- Improved system prompts to more clearly guide the model toward using dedicated tools (Read, Edit, Glob, Grep) instead of bash equivalents (`cat`, `sed`, `grep`, `find`), reducing unnecessary bash command usage
239- Improved PDF and request size error messages to show actual limits (100 pages, 20MB)
240- Reduced layout jitter in the terminal when the spinner appears and disappears during streaming
241- Removed misleading Anthropic API pricing from model selector for third-party provider (Bedrock, Vertex, Foundry) users
242
243## 2.1.30
244
245- Added `pages` parameter to the Read tool for PDFs, allowing specific page ranges to be read (e.g., `pages: "1-5"`). Large PDFs (>10 pages) now return a lightweight reference when `@` mentioned instead of being inlined into context.
246- Added pre-configured OAuth client credentials for MCP servers that don't support Dynamic Client Registration (e.g., Slack). Use `--client-id` and `--client-secret` with `claude mcp add`.
247- Added `/debug` for Claude to help troubleshoot the current session
248- Added support for additional `git log` and `git show` flags in read-only mode (e.g., `--topo-order`, `--cherry-pick`, `--format`, `--raw`)
249- Added token count, tool uses, and duration metrics to Task tool results
250- Added reduced motion mode to the config
251- Fixed phantom "(no content)" text blocks appearing in API conversation history, reducing token waste and potential model confusion
252- Fixed prompt cache not correctly invalidating when tool descriptions or input schemas changed, only when tool names changed
253- Fixed 400 errors that could occur after running `/login` when the conversation contained thinking blocks
254- Fixed a hang when resuming sessions with corrupted transcript files containing `parentUuid` cycles
255- Fixed rate limit message showing incorrect "/upgrade" suggestion for Max 20x users when extra-usage is unavailable
256- Fixed permission dialogs stealing focus while actively typing
257- Fixed subagents not being able to access SDK-provided MCP tools because they were not synced to the shared application state
258- Fixed a regression where Windows users with a `.bashrc` file could not run bash commands
259- Improved memory usage for `--resume` (68% reduction for users with many sessions) by replacing the session index with lightweight stat-based loading and progressive enrichment
260- Improved `TaskStop` tool to display the stopped command/task description in the result line instead of a generic "Task stopped" message
261- Changed `/model` to execute immediately instead of being queued
262- [VSCode] Added multiline input support to the "Other" text input in question dialogs (use Shift+Enter for new lines)
263- [VSCode] Fixed duplicate sessions appearing in the session list when starting a new conversation
264
265## 2.1.29
266
267- Fixed startup performance issues when resuming sessions that have `saved_hook_context`
268
269## 2.1.27
270
271- Added tool call failures and denials to debug logs
272- Fixed context management validation error for gateway users, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
273- Added `--from-pr` flag to resume sessions linked to a specific GitHub PR number or URL
274- Sessions are now automatically linked to PRs when created via `gh pr create`
275- Fixed /context command not displaying colored output
276- Fixed status bar duplicating background task indicator when PR status was shown
277- Windows: Fixed bash command execution failing for users with `.bashrc` files
278- Windows: Fixed console windows flashing when spawning child processes
279- VSCode: Fixed OAuth token expiration causing 401 errors after extended sessions
280
281## 2.1.25
282
283- Fixed beta header validation error for gateway users on Bedrock and Vertex, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
284
285## 2.1.23
286
287- Added customizable spinner verbs setting (`spinnerVerbs`)
288- Fixed mTLS and proxy connectivity for users behind corporate proxies or using client certificates
289- Fixed per-user temp directory isolation to prevent permission conflicts on shared systems
290- Fixed a race condition that could cause 400 errors when prompt caching scope was enabled
291- Fixed pending async hooks not being cancelled when headless streaming sessions ended
292- Fixed tab completion not updating the input field when accepting a suggestion
293- Fixed ripgrep search timeouts silently returning empty results instead of reporting errors
294- Improved terminal rendering performance with optimized screen data layout
295- Changed Bash commands to show timeout duration alongside elapsed time
296- Changed merged pull requests to show a purple status indicator in the prompt footer
297- [IDE] Fixed model options displaying incorrect region strings for Bedrock users in headless mode
298
299## 2.1.22
300
301- Fixed structured outputs for non-interactive (-p) mode
302
303## 2.1.21
304
305- Added support for full-width (zenkaku) number input from Japanese IME in option selection prompts
306- Fixed shell completion cache files being truncated on exit
307- Fixed API errors when resuming sessions that were interrupted during tool execution
308- Fixed auto-compact triggering too early on models with large output token limits
309- Fixed task IDs potentially being reused after deletion
310- Fixed file search not working in VS Code extension on Windows
311- Improved read/search progress indicators to show "Reading…" while in progress and "Read" when complete
312- Improved Claude to prefer file operation tools (Read, Edit, Write) over bash equivalents (cat, sed, awk)
313- [VSCode] Added automatic Python virtual environment activation, ensuring `python` and `pip` commands use the correct interpreter (configurable via `claudeCode.usePythonEnvironment` setting)
314- [VSCode] Fixed message action buttons having incorrect background colors
315
316## 2.1.20
317
318- Added arrow key history navigation in vim normal mode when cursor cannot move further
319- Added external editor shortcut (Ctrl+G) to the help menu for better discoverability
320- Added PR review status indicator to the prompt footer, showing the current branch's PR state (approved, changes requested, pending, or draft) as a colored dot with a clickable link
321- Added support for loading `CLAUDE.md` files from additional directories specified via `--add-dir` flag (requires setting `CLAUDE_CODE_ADDITIONAL_DIRECTORIES_CLAUDE_MD=1`)
322- Added ability to delete tasks via the `TaskUpdate` tool
323- Fixed session compaction issues that could cause resume to load full history instead of the compact summary
324- Fixed agents sometimes ignoring user messages sent while actively working on a task
325- Fixed wide character (emoji, CJK) rendering artifacts where trailing columns were not cleared when replaced by narrower characters
326- Fixed JSON parsing errors when MCP tool responses contain special Unicode characters
327- Fixed up/down arrow keys in multi-line and wrapped text input to prioritize cursor movement over history navigation
328- Fixed draft prompt being lost when pressing UP arrow to navigate command history
329- Fixed ghost text flickering when typing slash commands mid-input
330- Fixed marketplace source removal not properly deleting settings
331- Fixed duplicate output in some commands like `/context`
332- Fixed task list sometimes showing outside the main conversation view
333- Fixed syntax highlighting for diffs occurring within multiline constructs like Python docstrings
334- Fixed crashes when cancelling tool use
335- Improved `/sandbox` command UI to show dependency status with installation instructions when dependencies are missing
336- Improved thinking status text with a subtle shimmer animation
337- Improved task list to dynamically adjust visible items based on terminal height
338- Improved fork conversation hint to show how to resume the original session
339- Changed collapsed read/search groups to show present tense ("Reading", "Searching for") while in progress, and past tense ("Read", "Searched for") when complete
340- Changed `ToolSearch` results to appear as a brief notification instead of inline in the conversation
341- Changed the `/commit-push-pr` skill to automatically post PR URLs to Slack channels when configured via MCP tools
342- Changed the `/copy` command to be available to all users
343- Changed background agents to prompt for tool permissions before launching
344- Changed permission rules like `Bash(*)` to be accepted and treated as equivalent to `Bash`
345- Changed config backups to be timestamped and rotated (keeping 5 most recent) to prevent data loss
346
347## 2.1.19
348
349- Added env var `CLAUDE_CODE_ENABLE_TASKS`, set to `false` to keep the old system temporarily
350- Added shorthand `$0`, `$1`, etc. for accessing individual arguments in custom commands
351- Fixed crashes on processors without AVX instruction support
352- Fixed dangling Claude Code processes when terminal is closed by catching EIO errors from `process.exit()` and using SIGKILL as fallback
353- Fixed `/rename` and `/tag` not updating the correct session when resuming from a different directory (e.g., git worktrees)
354- Fixed resuming sessions by custom title not working when run from a different directory
355- Fixed pasted text content being lost when using prompt stash (Ctrl+S) and restore
356- Fixed agent list displaying "Sonnet (default)" instead of "Inherit (default)" for agents without an explicit model setting
357- Fixed backgrounded hook commands not returning early, potentially causing the session to wait on a process that was intentionally backgrounded
358- Fixed file write preview omitting empty lines
359- Changed skills without additional permissions or hooks to be allowed without requiring approval
360- Changed indexed argument syntax from `$ARGUMENTS.0` to `$ARGUMENTS[0]` (bracket syntax)
361- [SDK] Added replay of `queued_command` attachment messages as `SDKUserMessageReplay` events when `replayUserMessages` is enabled
362- [VSCode] Enabled session forking and rewind functionality for all users
363
364## 2.1.18
365
366- Added customizable keyboard shortcuts. Configure keybindings per context, create chord sequences, and personalize your workflow. Run `/keybindings` to get started. Learn more at https://code.claude.com/docs/en/keybindings
367
368## 2.1.17
369
370- Fixed crashes on processors without AVX instruction support
371
372## 2.1.16
373
374- Added new task management system, including new capabilities like dependency tracking
375- [VSCode] Added native plugin management support
376- [VSCode] Added ability for OAuth users to browse and resume remote Claude sessions from the Sessions dialog
377- Fixed out-of-memory crashes when resuming sessions with heavy subagent usage
378- Fixed an issue where the "context remaining" warning was not hidden after running `/compact`
379- Fixed session titles on the resume screen not respecting the user's language setting
380- [IDE] Fixed a race condition on Windows where the Claude Code sidebar view container would not appear on start
381
382## 2.1.15
383
384- Added deprecation notification for npm installations - run `claude install` or see https://docs.anthropic.com/en/docs/claude-code/getting-started for more options
385- Improved UI rendering performance with React Compiler
386- Fixed the "Context left until auto-compact" warning not disappearing after running `/compact`
387- Fixed MCP stdio server timeout not killing child process, which could cause UI freezes
388
389## 2.1.14
390
391- Added history-based autocomplete in bash mode (`!`) - type a partial command and press Tab to complete from your bash command history
392- Added search to installed plugins list - type to filter by name or description
393- Added support for pinning plugins to specific git commit SHAs, allowing marketplace entries to install exact versions
394- Fixed a regression where the context window blocking limit was calculated too aggressively, blocking users at ~65% context usage instead of the intended ~98%
395- Fixed memory issues that could cause crashes when running parallel subagents
396- Fixed memory leak in long-running sessions where stream resources were not cleaned up after shell commands completed
397- Fixed `@` symbol incorrectly triggering file autocomplete suggestions in bash mode
398- Fixed `@`-mention menu folder click behavior to navigate into directories instead of selecting them
399- Fixed `/feedback` command generating invalid GitHub issue URLs when description is very long
400- Fixed `/context` command to show the same token count and percentage as the status line in verbose mode
401- Fixed an issue where `/config`, `/context`, `/model`, and `/todos` command overlays could close unexpectedly
402- Fixed slash command autocomplete selecting wrong command when typing similar commands (e.g., `/context` vs `/compact`)
403- Fixed inconsistent back navigation in plugin marketplace when only one marketplace is configured
404- Fixed iTerm2 progress bar not clearing properly on exit, preventing lingering indicators and bell sounds
405- Improved backspace to delete pasted text as a single token instead of one character at a time
406- [VSCode] Added `/usage` command to display current plan usage
407
408## 2.1.12
409
410- Fixed message rendering bug
411
412## 2.1.11
413
414- Fixed excessive MCP connection requests for HTTP/SSE transports
415
416## 2.1.10
417
418- Added new `Setup` hook event that can be triggered via `--init`, `--init-only`, or `--maintenance` CLI flags for repository setup and maintenance operations
419- Added keyboard shortcut 'c' to copy OAuth URL when browser doesn't open automatically during login
420- Fixed a crash when running bash commands containing heredocs with JavaScript template literals like `${index + 1}`
421- Improved startup to capture keystrokes typed before the REPL is fully ready
422- Improved file suggestions to show as removable attachments instead of inserting text when accepted
423- [VSCode] Added install count display to plugin listings
424- [VSCode] Added trust warning when installing plugins
425
426## 2.1.9
427
428- Added `auto:N` syntax for configuring the MCP tool search auto-enable threshold, where N is the context window percentage (0-100)
429- Added `plansDirectory` setting to customize where plan files are stored
430- Added external editor support (Ctrl+G) in AskUserQuestion "Other" input field
431- Added session URL attribution to commits and PRs created from web sessions
432- Added support for `PreToolUse` hooks to return `additionalContext` to the model
433- Added `${CLAUDE_SESSION_ID}` string substitution for skills to access the current session ID
434- Fixed long sessions with parallel tool calls failing with an API error about orphan tool_result blocks
435- Fixed MCP server reconnection hanging when cached connection promise never resolves
436- Fixed Ctrl+Z suspend not working in terminals using Kitty keyboard protocol (Ghostty, iTerm2, kitty, WezTerm)
437
438## 2.1.7
439
440- Added `showTurnDuration` setting to hide turn duration messages (e.g., "Cooked for 1m 6s")
441- Added ability to provide feedback when accepting permission prompts
442- Added inline display of agent's final response in task notifications, making it easier to see results without reading the full transcript file
443- Fixed security vulnerability where wildcard permission rules could match compound commands containing shell operators
444- Fixed false "file modified" errors on Windows when cloud sync tools, antivirus scanners, or Git touch file timestamps without changing content
445- Fixed orphaned tool_result errors when sibling tools fail during streaming execution
446- Fixed context window blocking limit being calculated using the full context window instead of the effective context window (which reserves space for max output tokens)
447- Fixed spinner briefly flashing when running local slash commands like `/model` or `/theme`
448- Fixed terminal title animation jitter by using fixed-width braille characters
449- Fixed plugins with git submodules not being fully initialized when installed
450- Fixed bash commands failing on Windows when temp directory paths contained characters like `t` or `n` that were misinterpreted as escape sequences
451- Improved typing responsiveness by reducing memory allocation overhead in terminal rendering
452- Enabled MCP tool search auto mode by default for all users. When MCP tool descriptions exceed 10% of the context window, they are automatically deferred and discovered via the MCPSearch tool instead of being loaded upfront. This reduces context usage for users with many MCP tools configured. Users can disable this by adding `MCPSearch` to `disallowedTools` in their settings.
453- Changed OAuth and API Console URLs from console.anthropic.com to platform.claude.com
454- [VSCode] Fixed `claudeProcessWrapper` setting passing the wrapper path instead of the Claude binary path
455
456## 2.1.6
457
458- Added search functionality to `/config` command for quickly filtering settings
459- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
460- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
461- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
462- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
463- Added an error display when the editor fails during Ctrl+G
464- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
465- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
466- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
467- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
468- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
469- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
470- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
471- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
472- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
473- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
474- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
475- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
476- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
477- Improved @ autocomplete with icons for different suggestion types and single-line formatting
478- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
479- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
480- Changed terminal title to "Claude Code" on startup for better window identification
481- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
482- [VSCode] Fixed usage indicator not updating after manual compact
483
484## 2.1.5
485
486- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
487
488## 2.1.4
489
490- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
491- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
492
493## 2.1.3
494
495- Merged slash commands and skills, simplifying the mental model with no change in behavior
496- Added release channel (`stable` or `latest`) toggle to `/config`
497- Added detection and warnings for unreachable permission rules, with warnings in `/doctor` and after saving rules that include the source of each rule and actionable fix guidance
498- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
499- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
500- Fixed mismatch between background task count in status bar and items shown in tasks dialog
501- Fixed sub-agents using the wrong model during conversation compaction
502- Fixed web search in sub-agents using incorrect model
503- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
504- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
505- Improved slash command suggestion readability by truncating long descriptions to 2 lines
506- Changed tool hook execution timeout from 60 seconds to 10 minutes
507- [VSCode] Added clickable destination selector for permission requests, allowing you to choose where settings are saved (this project, all projects, shared with team, or session only)
508
509## 2.1.2
510
511- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
512- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
513- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
514- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
515- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
516- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
517- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
518- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
519- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
520- Fixed updates incorrectly claiming another installation is in progress
521- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
522- Fixed remote session URL and teleport being broken when using `/tasks` command
523- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
524- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
525- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
526- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
527- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
528- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
529- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
530- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
531- [SDK] Changed minimum zod peer dependency to ^4.0.0
532- [VSCode] Fixed usage display not updating after manual compact
533
534## 2.1.0
535
536- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
537- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
538- Added support for `agent` field in skills to specify agent type for execution
539- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
540- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
541- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
542- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
543- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
544- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
545- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
546- Fixed Esc key with queued prompts to only move them to input without canceling the running task
547- Reduced permission prompts for complex bash commands
548- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
549- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
550- Fixed plugin path resolution for file-based marketplace sources
551- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
552- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
553- Fixed Claude in Chrome support for WSL environments
554- Fixed Windows native installer silently failing when executable creation fails
555- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
556- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
557- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
558- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
559- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
560- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
561- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
562- Added hooks support for skill and slash command frontmatter
563- Added new Vim motions: `;` and `,` to repeat f/F/t/T motions, `y` operator for yank with `yy`/`Y`, `p`/`P` for paste, text objects (`iw`, `aw`, `iW`, `aW`, `i"`, `a"`, `i'`, `a'`, `i(`, `a(`, `i[`, `a[`, `i{`, `a{`), `>>` and `<<` for indent/dedent, and `J` to join lines
564- Added `/plan` command shortcut to enable plan mode directly from the prompt
565- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
566- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
567- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
568- Added support for `once: true` config for hooks
569- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
570- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
571- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
572- Added left/right arrow key navigation for cycling through tabs in dialogs
573- Added real-time thinking block display in Ctrl+O transcript mode
574- Added filepath to full output in background bash task details dialog
575- Added Skills as a separate category in the context visualization
576- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
577- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
578- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
579- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
580- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
581- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
582- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
583- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
584- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
585- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
586- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
587- Fixed commands with `$()` command substitution failing with parse errors
588- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
589- Fixed bash command prefix extraction to correctly identify subcommands after global options (e.g., `git -C /path log` now correctly matches `Bash(git log:*)` rules)
590- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
591- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
592- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
593- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
594- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
595- Fixed subagents sometimes not inheriting the parent's model by default
596- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
597- Fixed duplicate Bash commands appearing in permission request option labels
598- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
599- Fixed background task completion notifications to appear proactively with bullet point
600- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
601- Fixed cursor disappearing after dismissing permission dialogs
602- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
603- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
604- Fixed images being silently dropped when queueing messages while backgrounding a task
605- Fixed large pasted images failing with "Image was too large" error
606- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
607- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
608- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
609- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
610- Fixed spinner token counter to properly accumulate tokens from subagents during execution
611- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
612- Fixed race condition where LSP tool could return "no server available" during startup
613- Fixed feedback submission hanging indefinitely when network requests timeout
614- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
615- Fixed hook success message showing trailing colon when hook has no output
616- Multiple optimizations to improve startup performance
617- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
618- Improved performance when reading Jupyter notebooks with many cells
619- Improved reliability for piped input like `cat refactor.md | claude`
620- Improved reliability for AskQuestion tool
621- Improved sed in-place edit commands to render as file edits with diff preview
622- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
623- Improved compaction reliability
624- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
625- Improved skills to show progress while executing, displaying tool uses as they happen
626- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
627- Improved skill suggestions to prioritize recently and frequently used skills
628- Improved spinner feedback when waiting for the first response token
629- Improved token count display in spinner to include tokens from background agents
630- Improved incremental output for async agents to give the main thread more control and visibility
631- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
632- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
633- Improved macOS screenshot paste reliability with TIFF format support
634- Improved `/stats` output
635- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
636- Changed "Interrupted" message color from red to grey for a less alarming appearance
637- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
638- Removed underline styling from image reference links
639- [SDK] Changed minimum zod peer dependency to ^4.0.0
640- [VSCode] Added currently selected model name to the context menu
641- [VSCode] Added descriptive labels on auto-accept permission button (e.g., "Yes, allow npm for this project" instead of "Yes, and don't ask again")
642- [VSCode] Fixed paragraph breaks not rendering in markdown content
643- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
644- [Windows] Fixed issue with improper rendering
645
646## 2.0.76
647
648- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
649
650## 2.0.75
651
652- Minor bugfixes
653
654## 2.0.74
655
656- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
657- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
658- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
659- Added syntax highlighting info to theme picker
660- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
661- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
662- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
663- Fixed a potential crash when syntax highlighting isn't initialized correctly
664- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
665- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
666- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
667- [Windows] Fixed issue with improper rendering
668- [VSCode] Added gift tag pictogram for year-end promotion message
669
670## 2.0.73
671
672- Added clickable `[Image #N]` links that open attached images in the default viewer
673- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
674- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
675- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
676- Fixed slow input history cycling and race condition that could overwrite text after message submission
677- Improved `/theme` command to open theme picker directly
678- Improved theme picker UI
679- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
680- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
681
682## 2.0.72
683
684- Added Claude in Chrome (Beta) feature that works with the Chrome extension (https://claude.ai/chrome) to let you control your browser directly from Claude Code
685- Reduced terminal flickering
686- Added scannable QR code to mobile app tip for quick app downloads
687- Added loading indicator when resuming conversations for better feedback
688- Fixed `/context` command not respecting custom system prompts in non-interactive mode
689- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
690- Improved @ mention file suggestion speed (~3× faster in git repositories)
691- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
692- Improved settings validation errors to be more prominent
693- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
694
695## 2.0.71
696
697- Added /config toggle to enable/disable prompt suggestions
698- Added `/settings` as an alias for the `/config` command
699- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
700- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
701- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
702- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
703- New syntax highlighting engine for native build
704
705## 2.0.70
706
707- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
708- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
709- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
710- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
711- Fixed input being cleared when processing queued commands while the user was typing
712- Fixed prompt suggestions replacing typed input when pressing Tab
713- Fixed diff view not updating when terminal is resized
714- Improved memory usage by 3x for large conversations
715- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
716- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
717- Fix thinking mode toggle in /config not persisting correctly
718- Improve UI for file creation permission dialog
719
720## 2.0.69
721
722- Minor bugfixes
723
724## 2.0.68
725
726- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
727- Fixed a bug where disallowed MCP tools were visible to the model
728- Fixed an issue where steering messages could be lost while a subagent is working
729- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
730- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
731- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
732
733## 2.0.67
734
735- Thinking mode is now enabled by default for Opus 4.5
736- Thinking mode configuration has moved to /config
737- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
738- Show reason why autoupdater is disabled in `/doctor`
739- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
740- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
741- Fixed scroll position resetting after deleting a permission rule in `/permissions`
742- Fixed word deletion (opt+delete) and word navigation (opt+arrow) not working correctly with non-Latin text such as Cyrillic, Greek, Arabic, Hebrew, Thai, and Chinese
743- Fixed `claude install --force` not bypassing stale lock files
744- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
745- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
746
747## 2.0.65
748
749- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
750- Added context window information to status line input
751- Added `fileSuggestion` setting for custom `@` file search commands
752- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
753- Fixed prompt not being saved to history when aborting a query with Escape
754- Fixed Read tool image handling to identify format from bytes instead of file extension
755
756## 2.0.64
757
758- Made auto-compacting instant
759- Agents and bash commands can run asynchronously and send messages to wake up the main agent
760- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
761- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
762- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
763- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
764- Fixed auto-loading .env when using native installer
765- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
766- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
767- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
768- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
769- Bedrock: Improve efficiency of token counting
770- Bedrock: Add support for `aws login` AWS Management Console credentials
771- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
772
773## 2.0.62
774
775- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
776- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
777- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
778- Fixed slash command selection not working when multiple commands share the same name
779- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
780- Fixed running versions getting removed because lock file incorrectly going stale
781- Fixed IDE diff tab not closing when rejecting file changes
782
783## 2.0.61
784
785- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
786
787## 2.0.60
788
789- Added background agent support. Agents run in the background while you work
790- Added --disable-slash-commands CLI flag to disable all slash commands
791- Added model name to "Co-Authored-By" commit messages
792- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
793- Updated Fetch to skip summarization for pre-approved websites
794- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
795
796## 2.0.59
797
798- Added --agent CLI flag to override the agent setting for the current session
799- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
800- VS Code: Fixed .claude.json config file being read from incorrect location
801
802## 2.0.58
803
804- Pro users now have access to Opus 4.5 as part of their subscription!
805- Fixed timer duration showing "11m 60s" instead of "12m 0s"
806- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
807
808## 2.0.57
809
810- Added feedback input when rejecting plans, allowing users to tell Claude what to change
811- VSCode: Added streaming message support for real-time response display
812
813## 2.0.56
814
815- Added setting to enable/disable terminal progress bar (OSC 9;4)
816- VSCode Extension: Added support for VS Code's secondary sidebar (VS Code 1.97+), allowing Claude Code to be displayed in the right sidebar while keeping the file explorer on the left. Requires setting sidebar as Preferred Location in the config.
817
818## 2.0.55
819
820- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
821- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
822- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
823- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
824
825## 2.0.54
826
827- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
828- Fix issue with excessive iTerm notifications
829
830## 2.0.52
831
832- Fixed duplicate message display when starting Claude with a command line argument
833- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
834- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
835- Permit some uses of `$!` in bash commands
836
837## 2.0.51
838
839- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
840- Introducing Claude Code for Desktop: https://claude.com/download
841- To give you room to try out our new model, we've updated usage limits for Claude Code users. See the Claude Opus 4.5 blog for full details
842- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
843- Plan Mode now builds more precise plans and executes more thoroughly
844- Usage limit notifications now easier to understand
845- Switched `/usage` back to "% used"
846- Fixed handling of thinking errors
847- Fixed performance regression
848
849## 2.0.50
850
851- Fixed bug preventing calling MCP tools that have nested references in their input schemas
852- Silenced a noisy but harmless error during upgrades
853- Improved ultrathink text display
854- Improved clarity of 5-hour session limit warning message
855
856## 2.0.49
857
858- Added readline-style ctrl-y for pasting deleted text
859- Improved clarity of usage limit warning message
860- Fixed handling of subagent permissions
861
862## 2.0.47
863
864- Improved error messages and validation for `claude --teleport`
865- Improved error handling in `/usage`
866- Fixed race condition with history entry not getting logged at exit
867- Fixed Vertex AI configuration not being applied from `settings.json`
868
869## 2.0.46
870
871- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
872
873## 2.0.45
874
875- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
876- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
877- Send background tasks to Claude Code on the web by starting a message with `&`
878
879## 2.0.43
880
881- Added `permissionMode` field for custom agents
882- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
883- Added skills frontmatter field to declare skills to auto-load for subagents
884- Added the `SubagentStart` hook event
885- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
886- Fixed duplicate rendering of some messages in the UI
887- Fixed some visual flickers
888- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
889
890## 2.0.42
891
892- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
893
894## 2.0.41
895
896- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
897- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
898- Fixed incorrect labeling of user settings vs project settings in command descriptions
899- Fixed crash when plugin command hooks timeout during execution
900- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
901- Fixed broken security documentation links in trust dialogs and onboarding
902- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
903- ctrl-r history search landing on a slash command no longer cancels the search
904- SDK: Support custom timeouts for hooks
905- Allow more safe git commands to run without approval
906- Plugins: Added support for sharing and installing output styles
907- Teleporting a session from web will automatically set the upstream branch
908
909## 2.0.37
910
911- Fixed how idleness is computed for notifications
912- Hooks: Added matcher values for Notification hook events
913- Output Styles: Added `keep-coding-instructions` option to frontmatter
914
915## 2.0.36
916
917- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
918- Fixed queued messages being incorrectly executed as bash commands
919- Fixed input being lost when typing while a queued message is processed
920
921## 2.0.35
922
923- Improve fuzzy search results when searching commands
924- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
925- Added `CLAUDE_CODE_EXIT_AFTER_STOP_DELAY` environment variable to automatically exit SDK mode after a specified idle duration, useful for automated workflows and scripts
926- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
927- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
928
929## 2.0.34
930
931- VSCode Extension: Added setting to configure the initial permission mode for new conversations
932- Improved file path suggestion performance with native Rust-based fuzzy finder
933- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
934- Fixed memory crash when reading or writing large files (especially base64-encoded images)
935
936## 2.0.33
937
938- Native binary installs now launch quicker.
939- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
940- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
941
942## 2.0.32
943
944- Un-deprecate output styles based on community feedback
945- Added `companyAnnouncements` setting for displaying announcements on startup
946- Fixed hook progress messages not updating correctly during PostToolUse hook execution
947
948## 2.0.31
949
950- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
951- Vertex: add support for Web Search on supported models
952- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
953- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
954- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
955- Fixed plugin uninstall not removing plugins
956
957## 2.0.30
958
959- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
960- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
961- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
962- Added prompt-based stop hooks
963- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
964- Enabled SSE MCP servers on native build
965- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
966- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
967- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
968- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
969- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
970- Fixed bug that saved session permissions to local settings
971- Fixed MCP tools not being available to sub-agents
972- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
973- Fixed delay when navigating through typeahead suggestions with arrow keys
974- VSCode: Restored selection indicator in input footer showing current file or code selection status
975
976## 2.0.28
977
978- Plan mode: introduced new Plan subagent
979- Subagents: claude can now choose to resume subagents
980- Subagents: claude can dynamically choose the model used by its subagents
981- SDK: added --max-budget-usd flag
982- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
983- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
984- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
985- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
986- Various other bug fixes
987
988## 2.0.27
989
990- New UI for permission prompts
991- Added current branch filtering and search to session resume screen for easier navigation
992- Fixed directory @-mention causing "No assistant message found" error
993- VSCode Extension: Add config setting to include .gitignored files in file searches
994- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
995
996## 2.0.25
997
998- Removed legacy SDK entrypoint. Please migrate to @anthropic-ai/claude-agent-sdk for future SDK updates: https://platform.claude.com/docs/en/agent-sdk/migration-guide
999
1000## 2.0.24
1001
1002- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
1003- Claude Code Web: Support for Web -> CLI teleport
1004- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
1005- Bedrock: Display awsAuthRefresh output when auth is required
1006
1007## 2.0.22
1008
1009- Fixed content layout shift when scrolling through slash commands
1010- IDE: Add toggle to enable/disable thinking.
1011- Fix bug causing duplicate permission prompts with parallel tool calls
1012- Add support for enterprise managed MCP allowlist and denylist
1013
1014## 2.0.21
1015
1016- Support MCP `structuredContent` field in tool responses
1017- Added an interactive question tool
1018- Claude will now ask you questions more often in plan mode
1019- Added Haiku 4.5 as a model option for Pro users
1020- Fixed an issue where queued commands don't have access to previous messages' output
1021
1022## 2.0.20
1023
1024- Added support for Claude Skills
1025
1026## 2.0.19
1027
1028- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
1029- Fixed a bug where Haiku was unnecessarily called in print mode
1030
1031## 2.0.17
1032
1033- Added Haiku 4.5 to model selector!
1034- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
1035- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
1036- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
1037- OTEL: support HTTP_PROXY and HTTPS_PROXY
1038- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
1039
1040## 2.0.15
1041
1042- Fixed bug with resuming where previously created files needed to be read again before writing
1043- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
1044
1045## 2.0.14
1046
1047- Fix @-mentioning MCP servers to toggle them on/off
1048- Improve permission checks for bash with inline env vars
1049- Fix ultrathink + thinking toggle
1050- Reduce unnecessary logins
1051- Document --system-prompt
1052- Several improvements to rendering
1053- Plugins UI polish
1054
1055## 2.0.13
1056
1057- Fixed `/plugin` not working on native build
1058
1059## 2.0.12
1060
1061- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
1062- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
1063- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
1064- `/plugin validate` command for validating plugin structure and configuration
1065- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
1066- Plugin documentation available at https://code.claude.com/docs/en/plugins
1067- Comprehensive error messages and diagnostics via `/doctor` command
1068- Avoid flickering in `/model` selector
1069- Improvements to `/help`
1070- Avoid mentioning hooks in `/resume` summaries
1071- Changes to the "verbose" setting in `/config` now persist across sessions
1072
1073## 2.0.11
1074
1075- Reduced system prompt size by 1.4k tokens
1076- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
1077- Fixed Opus fallback rate limit errors appearing incorrectly
1078- Fixed /add-dir command selecting wrong default tab
1079
1080## 2.0.10
1081
1082- Rewrote terminal renderer for buttery smooth UI
1083- Enable/disable MCP servers by @mentioning, or in /mcp
1084- Added tab completion for shell commands in bash mode
1085- PreToolUse hooks can now modify tool inputs
1086- Press Ctrl-G to edit your prompt in your system's configured text editor
1087- Fixes for bash permission checks with environment variables in the command
1088
1089## 2.0.9
1090
1091- Fix regression where bash backgrounding stopped working
1092
1093## 2.0.8
1094
1095- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
1096- IDE: Add drag-and-drop support for files and folders in chat
1097- /context: Fix counting for thinking blocks
1098- Improve message rendering for users with light themes on dark terminals
1099- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
1100
1101## 2.0.5
1102
1103- IDE: Fix IME unintended message submission with Enter and Tab
1104- IDE: Add "Open in Terminal" link in login screen
1105- Fix unhandled OAuth expiration 401 API errors
1106- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
1107
1108## 2.0.1
1109
1110- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
1111- Various bug fixes and presentation improvements
1112
1113## 2.0.0
1114
1115- New native VS Code extension
1116- Fresh coat of paint throughout the whole app
1117- /rewind a conversation to undo code changes
1118- /usage command to see plan limits
1119- Tab to toggle thinking (sticky across sessions)
1120- Ctrl-R to search history
1121- Unshipped claude config command
1122- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
1123- SDK: The Claude Code SDK is now the Claude Agent SDK
1124- Add subagents dynamically with `--agents` flag
1125
1126## 1.0.126
1127
1128- Enable /context command for Bedrock and Vertex
1129- Add mTLS support for HTTP-based OpenTelemetry exporters
1130
1131## 1.0.124
1132
1133- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
1134- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
1135- No longer inform Claude of the list of allowed tools when permission is denied
1136- Fixed security vulnerability in Bash tool permission checks
1137- Improved VSCode extension performance for large files
1138
1139## 1.0.123
1140
1141- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
1142- Fixed thinking mode triggering on negation phrases like "don't think"
1143- Fixed rendering performance degradation during token streaming
1144- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
1145- Enhanced BashTool environment snapshot logging
1146- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
1147- Migrated --debug logging to a file, to enable easy tailing & filtering
1148
1149## 1.0.120
1150
1151- Fix input lag during typing, especially noticeable with large prompts
1152- Improved VSCode extension command registry and sessions dialog user experience
1153- Enhanced sessions dialog responsiveness and visual feedback
1154- Fixed IDE compatibility issue by removing worktree support check
1155- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
1156
1157## 1.0.119
1158
1159- Fix Windows issue where process visually freezes on entering interactive mode
1160- Support dynamic headers for MCP servers via headersHelper configuration
1161- Fix thinking mode not working in headless sessions
1162- Fix slash commands now properly update allowed tools instead of replacing them
1163
1164## 1.0.117
1165
1166- Add Ctrl-R history search to recall previous commands like bash/zsh
1167- Fix input lag while typing, especially on Windows
1168- Add sed command to auto-allowed commands in acceptEdits mode
1169- Fix Windows PATH comparison to be case-insensitive for drive letters
1170- Add permissions management hint to /add-dir output
1171
1172## 1.0.115
1173
1174- Improve thinking mode display with enhanced visual effects
1175- Type /t to temporarily disable thinking mode in your prompt
1176- Improve path validation for glob and grep tools
1177- Show condensed output for post-tool hooks to reduce visual clutter
1178- Fix visual feedback when loading state completes
1179- Improve UI consistency for permission request dialogs
1180
1181## 1.0.113
1182
1183- Deprecated piped input in interactive mode
1184- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
1185
1186## 1.0.112
1187
1188- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
1189- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
1190- Hooks: Added systemMessage support for SessionEnd hooks
1191- Added `spinnerTipsEnabled` setting to disable spinner tips
1192- IDE: Various improvements and bug fixes
1193
1194## 1.0.111
1195
1196- /model now validates provided model names
1197- Fixed Bash tool crashes caused by malformed shell syntax parsing
1198
1199## 1.0.110
1200
1201- /terminal-setup command now supports WezTerm
1202- MCP: OAuth tokens now proactively refresh before expiration
1203- Fixed reliability issues with background Bash processes
1204
1205## 1.0.109
1206
1207- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
1208
1209## 1.0.106
1210
1211- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
1212
1213## 1.0.97
1214
1215- Settings: /doctor now validates permission rule syntax and suggests corrections
1216
1217## 1.0.94
1218
1219- Vertex: add support for global endpoints for supported models
1220- /memory command now allows direct editing of all imported memory files
1221- SDK: Add custom tools as callbacks
1222- Added /todos command to list current todo items
1223
1224## 1.0.93
1225
1226- Windows: Add alt + v shortcut for pasting images from clipboard
1227- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
1228
1229## 1.0.90
1230
1231- Settings file changes take effect immediately - no restart required
1232
1233## 1.0.88
1234
1235- Fixed issue causing "OAuth authentication is currently not supported"
1236- Status line input now includes `exceeds_200k_tokens`
1237- Fixed incorrect usage tracking in /cost.
1238- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
1239- Bedrock: Updated default Sonnet model to Sonnet 4
1240
1241## 1.0.86
1242
1243- Added /context to help users self-serve debug context issues
1244- SDK: Added UUID support for all SDK messages
1245- SDK: Added `--replay-user-messages` to replay user messages back to stdout
1246
1247## 1.0.85
1248
1249- Status line input now includes session cost info
1250- Hooks: Introduced SessionEnd hook
1251
1252## 1.0.84
1253
1254- Fix tool_use/tool_result id mismatch error when network is unstable
1255- Fix Claude sometimes ignoring real-time steering when wrapping up a task
1256- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
1257- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
1258
1259## 1.0.83
1260
1261- @-mention: Support files with spaces in path
1262- New shimmering spinner
1263
1264## 1.0.82
1265
1266- SDK: Add request cancellation support
1267- SDK: New additionalDirectories option to search custom paths, improved slash command processing
1268- Settings: Validation prevents invalid fields in .claude/settings.json files
1269- MCP: Improve tool name consistency
1270- Bash: Fix crash when Claude tries to automatically read large files
1271
1272## 1.0.81
1273
1274- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
1275- Agents: Fix custom agent loading when agent files are unparsable
1276
1277## 1.0.80
1278
1279- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
1280
1281## 1.0.77
1282
1283- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
1284- SDK: Add session support and permission denial tracking
1285- Fix token limit errors in conversation summarization
1286- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
1287
1288## 1.0.73
1289
1290- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
1291- MCP: Press Esc to cancel OAuth authentication flows
1292- Bash: Improved command validation and reduced false security warnings
1293- UI: Enhanced spinner animations and status line visual hierarchy
1294- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
1295
1296## 1.0.72
1297
1298- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
1299
1300## 1.0.71
1301
1302- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
1303- Customizable status line: add your terminal prompt to Claude Code with /statusline
1304
1305## 1.0.70
1306
1307- Performance: Optimized message rendering for better performance with large contexts
1308- Windows: Fixed native file search, ripgrep, and subagent functionality
1309- Added support for @-mentions in slash command arguments
1310
1311## 1.0.69
1312
1313- Upgraded Opus to version 4.1
1314
1315## 1.0.68
1316
1317- Fix incorrect model names being used for certain commands like `/pr-comments`
1318- Windows: improve permissions checks for allow / deny tools and project trust. This may create a new project entry in `.claude.json` - manually merge the history field if desired.
1319- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
1320- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
1321- SDK: Added canUseTool callback support for tool confirmation
1322- Added `disableAllHooks` setting
1323- Improved file suggestions performance in large repos
1324
1325## 1.0.65
1326
1327- IDE: Fixed connection stability issues and error handling for diagnostics
1328- Windows: Fixed shell environment setup for users without .bashrc files
1329
1330## 1.0.64
1331
1332- Agents: Added model customization support - you can now specify which model an agent should use
1333- Agents: Fixed unintended access to the recursive agent tool
1334- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
1335- SDK: Fixed user input tracking across multi-turn conversations
1336- Added hidden files to file search and @-mention suggestions
1337
1338## 1.0.63
1339
1340- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
1341
1342## 1.0.62
1343
1344- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
1345- Hooks: Added SessionStart hook for new session initialization
1346- /add-dir command now supports typeahead for directory paths
1347- Improved network connectivity check reliability
1348
1349## 1.0.61
1350
1351- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
1352- Settings: Added `--settings` flag to load settings from a JSON file
1353- Settings: Fixed resolution of settings files paths that are symlinks
1354- OTEL: Fixed reporting of wrong organization after authentication changes
1355- Slash commands: Fixed permissions checking for allowed-tools with Bash
1356- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
1357- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
1358- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
1359
1360## 1.0.60
1361
1362- You can now create custom subagents for specialized tasks! Run /agents to get started
1363
1364## 1.0.59
1365
1366- SDK: Added tool confirmation support with canUseTool callback
1367- SDK: Allow specifying env for spawned process
1368- Hooks: Exposed PermissionDecision to hooks (including "ask")
1369- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
1370- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
1371
1372## 1.0.58
1373
1374- Added support for reading PDFs
1375- MCP: Improved server health status display in 'claude mcp list'
1376- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
1377
1378## 1.0.57
1379
1380- Added support for specifying a model in slash commands
1381- Improved permission messages to help Claude understand allowed tools
1382- Fix: Remove trailing newlines from bash output in terminal wrapping
1383
1384## 1.0.56
1385
1386- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
1387- Fixes for WSL IDE detection
1388- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
1389
1390## 1.0.55
1391
1392- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
1393- Windows: fixed Ctrl+Z crash
1394- SDK: Added ability to capture error logging
1395- Add --system-prompt-file option to override system prompt in print mode
1396
1397## 1.0.54
1398
1399- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
1400- Custom slash commands: Added argument-hint to frontmatter
1401- Windows: OAuth uses port 45454 and properly constructs browser URL
1402- Windows: mode switching now uses alt + m, and plan mode renders properly
1403- Shell: Switch to in-memory shell snapshot to fix file-related errors
1404
1405## 1.0.53
1406
1407- Updated @-mention file truncation from 100 lines to 2000 lines
1408- Add helper script settings for AWS token refresh: awsAuthRefresh (for foreground operations like aws sso login) and awsCredentialExport (for background operation with STS-like response).
1409
1410## 1.0.52
1411
1412- Added support for MCP server instructions
1413
1414## 1.0.51
1415
1416- Added support for native Windows (requires Git for Windows)
1417- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
1418- Settings: /doctor can now help you identify and fix invalid setting files
1419- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
1420- Increased auto-compact warning threshold from 60% to 80%
1421- Fixed an issue with handling user directories with spaces for shell snapshots
1422- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
1423- Custom slash commands: Fixed user-level commands in subdirectories
1424- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
1425
1426## 1.0.48
1427
1428- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
1429- Added progress messages to Bash tool based on the last 5 lines of command output
1430- Added expanding variables support for MCP server configuration
1431- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
1432- Improved IDE extension path handling when Claude Code runs in WSL
1433- Hooks: Added a PreCompact hook
1434- Vim mode: Added c, f/F, t/T
1435
1436## 1.0.45
1437
1438- Redesigned Search (Grep) tool with new tool input parameters and features
1439- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
1440- Fixed config file corruption issue by enforcing atomic writes
1441- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
1442- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
1443- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
1444
1445## 1.0.44
1446
1447- New /export command lets you quickly export a conversation for sharing
1448- MCP: resource_link tool results are now supported
1449- MCP: tool annotations and tool titles now display in /mcp view
1450- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
1451
1452## 1.0.43
1453
1454- Fixed a bug where the theme selector was saving excessively
1455- Hooks: Added EPIPE system error handling
1456
1457## 1.0.42
1458
1459- Added tilde (`~`) expansion support to `/add-dir` command
1460
1461## 1.0.41
1462
1463- Hooks: Split Stop hook triggering into Stop and SubagentStop
1464- Hooks: Enabled optional timeout configuration for each command
1465- Hooks: Added "hook_event_name" to hook input
1466- Fixed a bug where MCP tools would display twice in tool list
1467- New tool parameters JSON for Bash tool in `tool_decision` event
1468
1469## 1.0.40
1470
1471- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1472
1473## 1.0.39
1474
1475- New Active Time metric in OpenTelemetry logging
1476
1477## 1.0.38
1478
1479- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1480
1481## 1.0.37
1482
1483- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1484
1485## 1.0.36
1486
1487- Web search now takes today's date into context
1488- Fixed a bug where stdio MCP servers were not terminating properly on exit
1489
1490## 1.0.35
1491
1492- Added support for MCP OAuth Authorization Server discovery
1493
1494## 1.0.34
1495
1496- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1497
1498## 1.0.33
1499
1500- Improved logging functionality with session ID support
1501- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1502- Improvements to plan mode
1503
1504## 1.0.32
1505
1506- Updated loopback config for litellm
1507- Added forceLoginMethod setting to bypass login selection screen
1508
1509## 1.0.31
1510
1511- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1512
1513## 1.0.30
1514
1515- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1516- Improved file path autocomplete with filename matching
1517- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1518- Enhanced jq regex support for complex filters with pipes and select
1519
1520## 1.0.29
1521
1522- Improved CJK character support in cursor navigation and rendering
1523
1524## 1.0.28
1525
1526- Slash commands: Fix selector display during history navigation
1527- Resizes images before upload to prevent API size limit errors
1528- Added XDG_CONFIG_HOME support to configuration directory
1529- Performance optimizations for memory usage
1530- New attributes (terminal.type, language) in OpenTelemetry logging
1531
1532## 1.0.27
1533
1534- Streamable HTTP MCP servers are now supported
1535- Remote MCP servers (SSE and HTTP) now support OAuth
1536- MCP resources can now be @-mentioned
1537- /resume slash command to switch conversations within Claude Code
1538
1539## 1.0.25
1540
1541- Slash commands: moved "project" and "user" prefixes to descriptions
1542- Slash commands: improved reliability for command discovery
1543- Improved support for Ghostty
1544- Improved web search reliability
1545
1546## 1.0.24
1547
1548- Improved /mcp output
1549- Fixed a bug where settings arrays got overwritten instead of merged
1550
1551## 1.0.23
1552
1553- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1554- Released Python SDK: pip install claude-code-sdk to get started
1555
1556## 1.0.22
1557
1558- SDK: Renamed `total_cost` to `total_cost_usd`
1559
1560## 1.0.21
1561
1562- Improved editing of files with tab-based indentation
1563- Fix for tool_use without matching tool_result errors
1564- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1565
1566## 1.0.18
1567
1568- Added --add-dir CLI argument for specifying additional working directories
1569- Added streaming input support without require -p flag
1570- Improved startup performance and session storage performance
1571- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1572- Added detailed MCP server tools display (/mcp)
1573- MCP authentication and permission improvements
1574- Added auto-reconnection for MCP SSE connections on disconnect
1575- Fixed issue where pasted content was lost when dialogs appeared
1576
1577## 1.0.17
1578
1579- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1580- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1581- MCP server list UI improvements
1582- Update Claude Code process title to display "claude" instead of "node"
1583
1584## 1.0.11
1585
1586- Claude Code can now also be used with a Claude Pro subscription
1587- Added /upgrade for smoother switching to Claude Max plans
1588- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1589- Improved shell configuration error handling
1590- Improved todo list handling during compaction
1591
1592## 1.0.10
1593
1594- Added markdown table support
1595- Improved streaming performance
1596
1597## 1.0.8
1598
1599- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1600- Increased default otel interval from 1s -> 5s
1601- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1602- Fixed a regression where search tools unnecessarily asked for permissions
1603- Added support for triggering thinking non-English languages
1604- Improved compacting UI
1605
1606## 1.0.7
1607
1608- Renamed /allowed-tools -> /permissions
1609- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1610- Deprecated claude config commands in favor of editing settings.json
1611- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1612- Improved error handling for /install-github-app
1613- Bugfixes, UI polish, and tool reliability improvements
1614
1615## 1.0.6
1616
1617- Improved edit reliability for tab-indented files
1618- Respect CLAUDE_CONFIG_DIR everywhere
1619- Reduced unnecessary tool permission prompts
1620- Added support for symlinks in @file typeahead
1621- Bugfixes, UI polish, and tool reliability improvements
1622
1623## 1.0.4
1624
1625- Fixed a bug where MCP tool errors weren't being parsed correctly
1626
1627## 1.0.1
1628
1629- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1630- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1631- Updated documentation links and OAuth process descriptions
1632
1633## 1.0.0
1634
1635- Claude Code is now generally available
1636- Introducing Sonnet 4 and Opus 4 models
1637
1638## 0.2.125
1639
1640- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1641- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1642
1643## 0.2.117
1644
1645- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1646- Introduced settings.cleanupPeriodDays
1647- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1648- Introduced --debug mode
1649
1650## 0.2.108
1651
1652- You can now send messages to Claude while it works to steer Claude in real-time
1653- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1654- Fixed a bug where thinking was not working in -p mode
1655- Fixed a regression in /cost reporting
1656- Deprecated MCP wizard interface in favor of other MCP commands
1657- Lots of other bugfixes and improvements
1658
1659## 0.2.107
1660
1661- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1662
1663## 0.2.106
1664
1665- MCP SSE server configs can now specify custom headers
1666- Fixed a bug where MCP permission prompt didn't always show correctly
1667
1668## 0.2.105
1669
1670- Claude can now search the web
1671- Moved system & account status to /status
1672- Added word movement keybindings for Vim
1673- Improved latency for startup, todo tool, and file edits
1674
1675## 0.2.102
1676
1677- Improved thinking triggering reliability
1678- Improved @mention reliability for images and folders
1679- You can now paste multiple large chunks into one prompt
1680
1681## 0.2.100
1682
1683- Fixed a crash caused by a stack overflow error
1684- Made db storage optional; missing db support disables --continue and --resume
1685
1686## 0.2.98
1687
1688- Fixed an issue where auto-compact was running twice
1689
1690## 0.2.96
1691
1692- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1693
1694## 0.2.93
1695
1696- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1697- Claude now has access to a Todo list that helps it stay on track and be more organized
1698
1699## 0.2.82
1700
1701- Added support for --disallowedTools
1702- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1703
1704## 0.2.75
1705
1706- Hit Enter to queue up additional messages while Claude is working
1707- Drag in or copy/paste image files directly into the prompt
1708- @-mention files to directly add them to context
1709- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1710- Improved performance for filename auto-complete
1711
1712## 0.2.74
1713
1714- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1715- Task tool can now perform writes and run bash commands
1716
1717## 0.2.72
1718
1719- Updated spinner to indicate tokens loaded and tool usage
1720
1721## 0.2.70
1722
1723- Network commands like curl are now available for Claude to use
1724- Claude can now run multiple web queries in parallel
1725- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1726
1727## 0.2.69
1728
1729- Fixed UI glitches with improved Select component behavior
1730- Enhanced terminal output display with better text truncation logic
1731
1732## 0.2.67
1733
1734- Shared project permission rules can be saved in .claude/settings.json
1735
1736## 0.2.66
1737
1738- Print mode (-p) now supports streaming output via --output-format=stream-json
1739- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1740
1741## 0.2.63
1742
1743- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1744
1745## 0.2.61
1746
1747- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1748- Enhanced image detection for more reliable clipboard paste functionality
1749- Fixed an issue where ESC key could crash the conversation history selector
1750
1751## 0.2.59
1752
1753- Copy+paste images directly into your prompt
1754- Improved progress indicators for bash and fetch tools
1755- Bugfixes for non-interactive mode (-p)
1756
1757## 0.2.54
1758
1759- Quickly add to Memory by starting your message with '#'
1760- Press ctrl+r to see full output for long tool results
1761- Added support for MCP SSE transport
1762
1763## 0.2.53
1764
1765- New web fetch tool lets Claude view URLs that you paste in
1766- Fixed a bug with JPEG detection
1767
1768## 0.2.50
1769
1770- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1771
1772## 0.2.49
1773
1774- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1775
1776## 0.2.47
1777
1778- Press Tab to auto-complete file and folder names
1779- Press Shift + Tab to toggle auto-accept for file edits
1780- Automatic conversation compaction for infinite conversation length (toggle with /config)
1781
1782## 0.2.44
1783
1784- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1785
1786## 0.2.41
1787
1788- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1789- MCP server startup no longer blocks the app from starting up
1790
1791## 0.2.37
1792
1793- New /release-notes command lets you view release notes at any time
1794- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1795
1796## 0.2.36
1797
1798- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1799- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1800
1801## 0.2.34
1802
1803- Vim bindings for text input - enable with /vim or /config
1804
1805## 0.2.32
1806
1807- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1808- Fix for some PersistentShell issues
1809
1810## 0.2.31
1811
1812- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1813- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1814
1815## 0.2.30
1816
1817- Added ANSI color theme for better terminal compatibility
1818- Fixed issue where slash command arguments weren't being sent properly
1819- (Mac-only) API keys are now stored in macOS Keychain
1820
1821## 0.2.26
1822
1823- New /approved-tools command for managing tool permissions
1824- Word-level diff display for improved code readability
1825- Fuzzy matching for slash commands
1826
1827## 0.2.21
1828
1829- Fuzzy matching for /commands