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 1fe9e36 · on Nov 7, 2025 History
918 lines · 34.9 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.36
4
5- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
6- Fixed queued messages being incorrectly executed as bash commands
7- Fixed input being lost when typing while a queued message is processed
8
9## 2.0.35
10
11- Improve fuzzy search results when searching commands
12- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
13- 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
14- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
15- Fixed messages returning null `stop_reason` and `stop_sequence` values
16- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
17
18## 2.0.34
19
20- VSCode Extension: Added setting to configure the initial permission mode for new conversations
21- Improved file path suggestion performance with native Rust-based fuzzy finder
22- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
23- Fixed memory crash when reading or writing large files (especially base64-encoded images)
24
25## 2.0.33
26
27- Native binary installs now launch quicker.
28- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
29- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
30
31## 2.0.32
32
33- Un-deprecate output styles based on community feedback
34- Added `companyAnnouncements` setting for displaying announcements on startup
35- Fixed hook progress messages not updating correctly during PostToolUse hook execution
36
37## 2.0.31
38
39- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
40- Vertex: add support for Web Search on supported models
41- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
42- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
43- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
44- Fixed plugin uninstall not removing plugins
45
46## 2.0.30
47
48- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
49- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
50- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
51- Added prompt-based stop hooks
52- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
53- Enabled SSE MCP servers on native build
54- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
55- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
56- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
57- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
58- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
59- Fixed bug that saved session permissions to local settings
60- Fixed MCP tools not being available to sub-agents
61- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
62- Fixed delay when navigating through typeahead suggestions with arrow keys
63- VSCode: Restored selection indicator in input footer showing current file or code selection status
64
65## 2.0.28
66
67- Plan mode: introduced new Plan subagent
68- Subagents: claude can now choose to resume subagents
69- Subagents: claude can dynamically choose the model used by its subagents
70- SDK: added --max-budget-usd flag
71- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
72- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
73- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
74- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
75- Various other bug fixes
76
77## 2.0.27
78
79- New UI for permission prompts
80- Added current branch filtering and search to session resume screen for easier navigation
81- Fixed directory @-mention causing "No assistant message found" error
82- VSCode Extension: Add config setting to include .gitignored files in file searches
83- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
84
85## 2.0.25
86
87- Removed legacy SDK entrypoint. Please migrate to @anthropic-ai/claude-agent-sdk for future SDK updates: https://docs.claude.com/en/docs/claude-code/sdk/migration-guide
88
89## 2.0.24
90
91- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
92- Claude Code Web: Support for Web -> CLI teleport
93- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
94- Bedrock: Display awsAuthRefresh output when auth is required
95
96## 2.0.22
97
98- Fixed content layout shift when scrolling through slash commands
99- IDE: Add toggle to enable/disable thinking.
100- Fix bug causing duplicate permission prompts with parallel tool calls
101- Add support for enterprise managed MCP allowlist and denylist
102
103## 2.0.21
104
105- Support MCP `structuredContent` field in tool responses
106- Added an interactive question tool
107- Claude will now ask you questions more often in plan mode
108- Added Haiku 4.5 as a model option for Pro users
109- Fixed an issue where queued commands don't have access to previous messages' output
110
111## 2.0.20
112
113- Added support for Claude Skills
114
115## 2.0.19
116
117- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
118- Fixed a bug where Haiku was unnecessarily called in print mode
119
120## 2.0.17
121
122- Added Haiku 4.5 to model selector!
123- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
124- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
125- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
126- OTEL: support HTTP_PROXY and HTTPS_PROXY
127- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
128
129## 2.0.15
130
131- Fixed bug with resuming where previously created files needed to be read again before writing
132- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
133
134## 2.0.14
135
136- Fix @-mentioning MCP servers to toggle them on/off
137- Improve permission checks for bash with inline env vars
138- Fix ultrathink + thinking toggle
139- Reduce unnecessary logins
140- Document --system-prompt
141- Several improvements to rendering
142- Plugins UI polish
143
144## 2.0.13
145
146- Fixed `/plugin` not working on native build
147
148## 2.0.12
149
150- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
151- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
152- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
153- `/plugin validate` command for validating plugin structure and configuration
154- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
155- Plugin documentation available at https://docs.claude.com/en/docs/claude-code/plugins
156- Comprehensive error messages and diagnostics via `/doctor` command
157- Avoid flickering in `/model` selector
158- Improvements to `/help`
159- Avoid mentioning hooks in `/resume` summaries
160- Changes to the "verbose" setting in `/config` now persist across sessions
161
162## 2.0.11
163
164- Reduced system prompt size by 1.4k tokens
165- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
166- Fixed Opus fallback rate limit errors appearing incorrectly
167- Fixed /add-dir command selecting wrong default tab
168
169## 2.0.10
170
171- Rewrote terminal renderer for buttery smooth UI
172- Enable/disable MCP servers by @mentioning, or in /mcp
173- Added tab completion for shell commands in bash mode
174- PreToolUse hooks can now modify tool inputs
175- Press Ctrl-G to edit your prompt in your system's configured text editor
176- Fixes for bash permission checks with environment variables in the command
177
178## 2.0.9
179
180- Fix regression where bash backgrounding stopped working
181
182## 2.0.8
183
184- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
185- IDE: Add drag-and-drop support for files and folders in chat
186- /context: Fix counting for thinking blocks
187- Improve message rendering for users with light themes on dark terminals
188- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
189
190## 2.0.5
191
192- IDE: Fix IME unintended message submission with Enter and Tab
193- IDE: Add "Open in Terminal" link in login screen
194- Fix unhandled OAuth expiration 401 API errors
195- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
196
197## 2.0.1
198
199- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
200- Various bug fixes and presentation improvements
201
202## 2.0.0
203
204- New native VS Code extension
205- Fresh coat of paint throughout the whole app
206- /rewind a conversation to undo code changes
207- /usage command to see plan limits
208- Tab to toggle thinking (sticky across sessions)
209- Ctrl-R to search history
210- Unshipped claude config command
211- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
212- SDK: The Claude Code SDK is now the Claude Agent SDK
213- Add subagents dynamically with `--agents` flag
214
215## 1.0.126
216
217- Enable /context command for Bedrock and Vertex
218- Add mTLS support for HTTP-based OpenTelemetry exporters
219
220## 1.0.124
221
222- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
223- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
224- No longer inform Claude of the list of allowed tools when permission is denied
225- Fixed security vulnerability in Bash tool permission checks
226- Improved VSCode extension performance for large files
227
228## 1.0.123
229
230- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
231- Fixed thinking mode triggering on negation phrases like "don't think"
232- Fixed rendering performance degradation during token streaming
233- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
234- Enhanced BashTool environment snapshot logging
235- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
236- Migrated --debug logging to a file, to enable easy tailing & filtering
237
238## 1.0.120
239
240- Fix input lag during typing, especially noticeable with large prompts
241- Improved VSCode extension command registry and sessions dialog user experience
242- Enhanced sessions dialog responsiveness and visual feedback
243- Fixed IDE compatibility issue by removing worktree support check
244- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
245
246## 1.0.119
247
248- Fix Windows issue where process visually freezes on entering interactive mode
249- Support dynamic headers for MCP servers via headersHelper configuration
250- Fix thinking mode not working in headless sessions
251- Fix slash commands now properly update allowed tools instead of replacing them
252
253## 1.0.117
254
255- Add Ctrl-R history search to recall previous commands like bash/zsh
256- Fix input lag while typing, especially on Windows
257- Add sed command to auto-allowed commands in acceptEdits mode
258- Fix Windows PATH comparison to be case-insensitive for drive letters
259- Add permissions management hint to /add-dir output
260
261## 1.0.115
262
263- Improve thinking mode display with enhanced visual effects
264- Type /t to temporarily disable thinking mode in your prompt
265- Improve path validation for glob and grep tools
266- Show condensed output for post-tool hooks to reduce visual clutter
267- Fix visual feedback when loading state completes
268- Improve UI consistency for permission request dialogs
269
270## 1.0.113
271
272- Deprecated piped input in interactive mode
273- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
274
275## 1.0.112
276
277- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
278- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
279- Hooks: Added systemMessage support for SessionEnd hooks
280- Added `spinnerTipsEnabled` setting to disable spinner tips
281- IDE: Various improvements and bug fixes
282
283## 1.0.111
284
285- /model now validates provided model names
286- Fixed Bash tool crashes caused by malformed shell syntax parsing
287
288## 1.0.110
289
290- /terminal-setup command now supports WezTerm
291- MCP: OAuth tokens now proactively refresh before expiration
292- Fixed reliability issues with background Bash processes
293
294## 1.0.109
295
296- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
297
298## 1.0.106
299
300- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
301
302## 1.0.97
303
304- Settings: /doctor now validates permission rule syntax and suggests corrections
305
306## 1.0.94
307
308- Vertex: add support for global endpoints for supported models
309- /memory command now allows direct editing of all imported memory files
310- SDK: Add custom tools as callbacks
311- Added /todos command to list current todo items
312
313## 1.0.93
314
315- Windows: Add alt + v shortcut for pasting images from clipboard
316- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
317
318## 1.0.90
319
320- Settings file changes take effect immediately - no restart required
321
322## 1.0.88
323
324- Fixed issue causing "OAuth authentication is currently not supported"
325- Status line input now includes `exceeds_200k_tokens`
326- Fixed incorrect usage tracking in /cost.
327- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
328- Bedrock: Updated default Sonnet model to Sonnet 4
329
330## 1.0.86
331
332- Added /context to help users self-serve debug context issues
333- SDK: Added UUID support for all SDK messages
334- SDK: Added `--replay-user-messages` to replay user messages back to stdout
335
336## 1.0.85
337
338- Status line input now includes session cost info
339- Hooks: Introduced SessionEnd hook
340
341## 1.0.84
342
343- Fix tool_use/tool_result id mismatch error when network is unstable
344- Fix Claude sometimes ignoring real-time steering when wrapping up a task
345- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
346- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
347
348## 1.0.83
349
350- @-mention: Support files with spaces in path
351- New shimmering spinner
352
353## 1.0.82
354
355- SDK: Add request cancellation support
356- SDK: New additionalDirectories option to search custom paths, improved slash command processing
357- Settings: Validation prevents invalid fields in .claude/settings.json files
358- MCP: Improve tool name consistency
359- Bash: Fix crash when Claude tries to automatically read large files
360
361## 1.0.81
362
363- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
364- Agents: Fix custom agent loading when agent files are unparsable
365
366## 1.0.80
367
368- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
369
370## 1.0.77
371
372- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
373- SDK: Add session support and permission denial tracking
374- Fix token limit errors in conversation summarization
375- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
376
377## 1.0.73
378
379- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
380- MCP: Press Esc to cancel OAuth authentication flows
381- Bash: Improved command validation and reduced false security warnings
382- UI: Enhanced spinner animations and status line visual hierarchy
383- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
384
385## 1.0.72
386
387- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
388
389## 1.0.71
390
391- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
392- Customizable status line: add your terminal prompt to Claude Code with /statusline
393
394## 1.0.70
395
396- Performance: Optimized message rendering for better performance with large contexts
397- Windows: Fixed native file search, ripgrep, and subagent functionality
398- Added support for @-mentions in slash command arguments
399
400## 1.0.69
401
402- Upgraded Opus to version 4.1
403
404## 1.0.68
405
406- Fix incorrect model names being used for certain commands like `/pr-comments`
407- 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.
408- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
409- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
410- SDK: Added canUseTool callback support for tool confirmation
411- Added `disableAllHooks` setting
412- Improved file suggestions performance in large repos
413
414## 1.0.65
415
416- IDE: Fixed connection stability issues and error handling for diagnostics
417- Windows: Fixed shell environment setup for users without .bashrc files
418
419## 1.0.64
420
421- Agents: Added model customization support - you can now specify which model an agent should use
422- Agents: Fixed unintended access to the recursive agent tool
423- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
424- SDK: Fixed user input tracking across multi-turn conversations
425- Added hidden files to file search and @-mention suggestions
426
427## 1.0.63
428
429- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
430
431## 1.0.62
432
433- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
434- Hooks: Added SessionStart hook for new session initialization
435- /add-dir command now supports typeahead for directory paths
436- Improved network connectivity check reliability
437
438## 1.0.61
439
440- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
441- Settings: Added `--settings` flag to load settings from a JSON file
442- Settings: Fixed resolution of settings files paths that are symlinks
443- OTEL: Fixed reporting of wrong organization after authentication changes
444- Slash commands: Fixed permissions checking for allowed-tools with Bash
445- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
446- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
447- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
448
449## 1.0.60
450
451- You can now create custom subagents for specialized tasks! Run /agents to get started
452
453## 1.0.59
454
455- SDK: Added tool confirmation support with canUseTool callback
456- SDK: Allow specifying env for spawned process
457- Hooks: Exposed PermissionDecision to hooks (including "ask")
458- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
459- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
460
461## 1.0.58
462
463- Added support for reading PDFs
464- MCP: Improved server health status display in 'claude mcp list'
465- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
466
467## 1.0.57
468
469- Added support for specifying a model in slash commands
470- Improved permission messages to help Claude understand allowed tools
471- Fix: Remove trailing newlines from bash output in terminal wrapping
472
473## 1.0.56
474
475- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
476- Fixes for WSL IDE detection
477- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
478
479## 1.0.55
480
481- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
482- Windows: fixed Ctrl+Z crash
483- SDK: Added ability to capture error logging
484- Add --system-prompt-file option to override system prompt in print mode
485
486## 1.0.54
487
488- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
489- Custom slash commands: Added argument-hint to frontmatter
490- Windows: OAuth uses port 45454 and properly constructs browser URL
491- Windows: mode switching now uses alt + m, and plan mode renders properly
492- Shell: Switch to in-memory shell snapshot to fix file-related errors
493
494## 1.0.53
495
496- Updated @-mention file truncation from 100 lines to 2000 lines
497- 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).
498
499## 1.0.52
500
501- Added support for MCP server instructions
502
503## 1.0.51
504
505- Added support for native Windows (requires Git for Windows)
506- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
507- Settings: /doctor can now help you identify and fix invalid setting files
508- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
509- Increased auto-compact warning threshold from 60% to 80%
510- Fixed an issue with handling user directories with spaces for shell snapshots
511- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
512- Custom slash commands: Fixed user-level commands in subdirectories
513- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
514
515## 1.0.48
516
517- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
518- Added progress messages to Bash tool based on the last 5 lines of command output
519- Added expanding variables support for MCP server configuration
520- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
521- Improved IDE extension path handling when Claude Code runs in WSL
522- Hooks: Added a PreCompact hook
523- Vim mode: Added c, f/F, t/T
524
525## 1.0.45
526
527- Redesigned Search (Grep) tool with new tool input parameters and features
528- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
529- Fixed config file corruption issue by enforcing atomic writes
530- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
531- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
532- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
533
534## 1.0.44
535
536- New /export command lets you quickly export a conversation for sharing
537- MCP: resource_link tool results are now supported
538- MCP: tool annotations and tool titles now display in /mcp view
539- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
540
541## 1.0.43
542
543- Fixed a bug where the theme selector was saving excessively
544- Hooks: Added EPIPE system error handling
545
546## 1.0.42
547
548- Added tilde (`~`) expansion support to `/add-dir` command
549
550## 1.0.41
551
552- Hooks: Split Stop hook triggering into Stop and SubagentStop
553- Hooks: Enabled optional timeout configuration for each command
554- Hooks: Added "hook_event_name" to hook input
555- Fixed a bug where MCP tools would display twice in tool list
556- New tool parameters JSON for Bash tool in `tool_decision` event
557
558## 1.0.40
559
560- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
561
562## 1.0.39
563
564- New Active Time metric in OpenTelemetry logging
565
566## 1.0.38
567
568- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://docs.claude.com/en/docs/claude-code/hooks
569
570## 1.0.37
571
572- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
573
574## 1.0.36
575
576- Web search now takes today's date into context
577- Fixed a bug where stdio MCP servers were not terminating properly on exit
578
579## 1.0.35
580
581- Added support for MCP OAuth Authorization Server discovery
582
583## 1.0.34
584
585- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
586
587## 1.0.33
588
589- Improved logging functionality with session ID support
590- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
591- Improvements to plan mode
592
593## 1.0.32
594
595- Updated loopback config for litellm
596- Added forceLoginMethod setting to bypass login selection screen
597
598## 1.0.31
599
600- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
601
602## 1.0.30
603
604- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
605- Improved file path autocomplete with filename matching
606- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
607- Enhanced jq regex support for complex filters with pipes and select
608
609## 1.0.29
610
611- Improved CJK character support in cursor navigation and rendering
612
613## 1.0.28
614
615- Slash commands: Fix selector display during history navigation
616- Resizes images before upload to prevent API size limit errors
617- Added XDG_CONFIG_HOME support to configuration directory
618- Performance optimizations for memory usage
619- New attributes (terminal.type, language) in OpenTelemetry logging
620
621## 1.0.27
622
623- Streamable HTTP MCP servers are now supported
624- Remote MCP servers (SSE and HTTP) now support OAuth
625- MCP resources can now be @-mentioned
626- /resume slash command to switch conversations within Claude Code
627
628## 1.0.25
629
630- Slash commands: moved "project" and "user" prefixes to descriptions
631- Slash commands: improved reliability for command discovery
632- Improved support for Ghostty
633- Improved web search reliability
634
635## 1.0.24
636
637- Improved /mcp output
638- Fixed a bug where settings arrays got overwritten instead of merged
639
640## 1.0.23
641
642- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
643- Released Python SDK: pip install claude-code-sdk to get started
644
645## 1.0.22
646
647- SDK: Renamed `total_cost` to `total_cost_usd`
648
649## 1.0.21
650
651- Improved editing of files with tab-based indentation
652- Fix for tool_use without matching tool_result errors
653- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
654
655## 1.0.18
656
657- Added --add-dir CLI argument for specifying additional working directories
658- Added streaming input support without require -p flag
659- Improved startup performance and session storage performance
660- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
661- Added detailed MCP server tools display (/mcp)
662- MCP authentication and permission improvements
663- Added auto-reconnection for MCP SSE connections on disconnect
664- Fixed issue where pasted content was lost when dialogs appeared
665
666## 1.0.17
667
668- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
669- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
670- MCP server list UI improvements
671- Update Claude Code process title to display "claude" instead of "node"
672
673## 1.0.11
674
675- Claude Code can now also be used with a Claude Pro subscription
676- Added /upgrade for smoother switching to Claude Max plans
677- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
678- Improved shell configuration error handling
679- Improved todo list handling during compaction
680
681## 1.0.10
682
683- Added markdown table support
684- Improved streaming performance
685
686## 1.0.8
687
688- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
689- Increased default otel interval from 1s -> 5s
690- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
691- Fixed a regression where search tools unnecessarily asked for permissions
692- Added support for triggering thinking non-English languages
693- Improved compacting UI
694
695## 1.0.7
696
697- Renamed /allowed-tools -> /permissions
698- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
699- Deprecated claude config commands in favor of editing settings.json
700- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
701- Improved error handling for /install-github-app
702- Bugfixes, UI polish, and tool reliability improvements
703
704## 1.0.6
705
706- Improved edit reliability for tab-indented files
707- Respect CLAUDE_CONFIG_DIR everywhere
708- Reduced unnecessary tool permission prompts
709- Added support for symlinks in @file typeahead
710- Bugfixes, UI polish, and tool reliability improvements
711
712## 1.0.4
713
714- Fixed a bug where MCP tool errors weren't being parsed correctly
715
716## 1.0.1
717
718- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
719- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
720- Updated documentation links and OAuth process descriptions
721
722## 1.0.0
723
724- Claude Code is now generally available
725- Introducing Sonnet 4 and Opus 4 models
726
727## 0.2.125
728
729- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
730- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
731
732## 0.2.117
733
734- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
735- Introduced settings.cleanupPeriodDays
736- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
737- Introduced --debug mode
738
739## 0.2.108
740
741- You can now send messages to Claude while it works to steer Claude in real-time
742- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
743- Fixed a bug where thinking was not working in -p mode
744- Fixed a regression in /cost reporting
745- Deprecated MCP wizard interface in favor of other MCP commands
746- Lots of other bugfixes and improvements
747
748## 0.2.107
749
750- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
751
752## 0.2.106
753
754- MCP SSE server configs can now specify custom headers
755- Fixed a bug where MCP permission prompt didn't always show correctly
756
757## 0.2.105
758
759- Claude can now search the web
760- Moved system & account status to /status
761- Added word movement keybindings for Vim
762- Improved latency for startup, todo tool, and file edits
763
764## 0.2.102
765
766- Improved thinking triggering reliability
767- Improved @mention reliability for images and folders
768- You can now paste multiple large chunks into one prompt
769
770## 0.2.100
771
772- Fixed a crash caused by a stack overflow error
773- Made db storage optional; missing db support disables --continue and --resume
774
775## 0.2.98
776
777- Fixed an issue where auto-compact was running twice
778
779## 0.2.96
780
781- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
782
783## 0.2.93
784
785- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
786- Claude now has access to a Todo list that helps it stay on track and be more organized
787
788## 0.2.82
789
790- Added support for --disallowedTools
791- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
792
793## 0.2.75
794
795- Hit Enter to queue up additional messages while Claude is working
796- Drag in or copy/paste image files directly into the prompt
797- @-mention files to directly add them to context
798- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
799- Improved performance for filename auto-complete
800
801## 0.2.74
802
803- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
804- Task tool can now perform writes and run bash commands
805
806## 0.2.72
807
808- Updated spinner to indicate tokens loaded and tool usage
809
810## 0.2.70
811
812- Network commands like curl are now available for Claude to use
813- Claude can now run multiple web queries in parallel
814- Pressing ESC once immediately interrupts Claude in Auto-accept mode
815
816## 0.2.69
817
818- Fixed UI glitches with improved Select component behavior
819- Enhanced terminal output display with better text truncation logic
820
821## 0.2.67
822
823- Shared project permission rules can be saved in .claude/settings.json
824
825## 0.2.66
826
827- Print mode (-p) now supports streaming output via --output-format=stream-json
828- Fixed issue where pasting could trigger memory or bash mode unexpectedly
829
830## 0.2.63
831
832- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
833
834## 0.2.61
835
836- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
837- Enhanced image detection for more reliable clipboard paste functionality
838- Fixed an issue where ESC key could crash the conversation history selector
839
840## 0.2.59
841
842- Copy+paste images directly into your prompt
843- Improved progress indicators for bash and fetch tools
844- Bugfixes for non-interactive mode (-p)
845
846## 0.2.54
847
848- Quickly add to Memory by starting your message with '#'
849- Press ctrl+r to see full output for long tool results
850- Added support for MCP SSE transport
851
852## 0.2.53
853
854- New web fetch tool lets Claude view URLs that you paste in
855- Fixed a bug with JPEG detection
856
857## 0.2.50
858
859- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
860
861## 0.2.49
862
863- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
864
865## 0.2.47
866
867- Press Tab to auto-complete file and folder names
868- Press Shift + Tab to toggle auto-accept for file edits
869- Automatic conversation compaction for infinite conversation length (toggle with /config)
870
871## 0.2.44
872
873- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
874
875## 0.2.41
876
877- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
878- MCP server startup no longer blocks the app from starting up
879
880## 0.2.37
881
882- New /release-notes command lets you view release notes at any time
883- `claude config add/remove` commands now accept multiple values separated by commas or spaces
884
885## 0.2.36
886
887- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
888- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
889
890## 0.2.34
891
892- Vim bindings for text input - enable with /vim or /config
893
894## 0.2.32
895
896- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
897- Fix for some PersistentShell issues
898
899## 0.2.31
900
901- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
902- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
903
904## 0.2.30
905
906- Added ANSI color theme for better terminal compatibility
907- Fixed issue where slash command arguments weren't being sent properly
908- (Mac-only) API keys are now stored in macOS Keychain
909
910## 0.2.26
911
912- New /approved-tools command for managing tool permissions
913- Word-level diff display for improved code readability
914- Fuzzy matching for slash commands
915
916## 0.2.21
917
918- Fuzzy matching for /commands