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 d38bde5 · on Nov 11, 2025 History
924 lines · 35.1 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.37
4
5- Fixed how idleness is computed for notifications
6- Hooks: Added matcher values for Notification hook events
7- Output Styles: Added `keep-coding-instructions` option to frontmatter
8
9## 2.0.36
10
11- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
12- Fixed queued messages being incorrectly executed as bash commands
13- Fixed input being lost when typing while a queued message is processed
14
15## 2.0.35
16
17- Improve fuzzy search results when searching commands
18- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
19- 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
20- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
21- Fixed messages returning null `stop_reason` and `stop_sequence` values
22- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
23
24## 2.0.34
25
26- VSCode Extension: Added setting to configure the initial permission mode for new conversations
27- Improved file path suggestion performance with native Rust-based fuzzy finder
28- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
29- Fixed memory crash when reading or writing large files (especially base64-encoded images)
30
31## 2.0.33
32
33- Native binary installs now launch quicker.
34- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
35- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
36
37## 2.0.32
38
39- Un-deprecate output styles based on community feedback
40- Added `companyAnnouncements` setting for displaying announcements on startup
41- Fixed hook progress messages not updating correctly during PostToolUse hook execution
42
43## 2.0.31
44
45- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
46- Vertex: add support for Web Search on supported models
47- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
48- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
49- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
50- Fixed plugin uninstall not removing plugins
51
52## 2.0.30
53
54- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
55- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
56- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
57- Added prompt-based stop hooks
58- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
59- Enabled SSE MCP servers on native build
60- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
61- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
62- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
63- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
64- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
65- Fixed bug that saved session permissions to local settings
66- Fixed MCP tools not being available to sub-agents
67- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
68- Fixed delay when navigating through typeahead suggestions with arrow keys
69- VSCode: Restored selection indicator in input footer showing current file or code selection status
70
71## 2.0.28
72
73- Plan mode: introduced new Plan subagent
74- Subagents: claude can now choose to resume subagents
75- Subagents: claude can dynamically choose the model used by its subagents
76- SDK: added --max-budget-usd flag
77- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
78- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
79- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
80- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
81- Various other bug fixes
82
83## 2.0.27
84
85- New UI for permission prompts
86- Added current branch filtering and search to session resume screen for easier navigation
87- Fixed directory @-mention causing "No assistant message found" error
88- VSCode Extension: Add config setting to include .gitignored files in file searches
89- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
90
91## 2.0.25
92
93- 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
94
95## 2.0.24
96
97- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
98- Claude Code Web: Support for Web -> CLI teleport
99- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
100- Bedrock: Display awsAuthRefresh output when auth is required
101
102## 2.0.22
103
104- Fixed content layout shift when scrolling through slash commands
105- IDE: Add toggle to enable/disable thinking.
106- Fix bug causing duplicate permission prompts with parallel tool calls
107- Add support for enterprise managed MCP allowlist and denylist
108
109## 2.0.21
110
111- Support MCP `structuredContent` field in tool responses
112- Added an interactive question tool
113- Claude will now ask you questions more often in plan mode
114- Added Haiku 4.5 as a model option for Pro users
115- Fixed an issue where queued commands don't have access to previous messages' output
116
117## 2.0.20
118
119- Added support for Claude Skills
120
121## 2.0.19
122
123- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
124- Fixed a bug where Haiku was unnecessarily called in print mode
125
126## 2.0.17
127
128- Added Haiku 4.5 to model selector!
129- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
130- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
131- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
132- OTEL: support HTTP_PROXY and HTTPS_PROXY
133- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
134
135## 2.0.15
136
137- Fixed bug with resuming where previously created files needed to be read again before writing
138- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
139
140## 2.0.14
141
142- Fix @-mentioning MCP servers to toggle them on/off
143- Improve permission checks for bash with inline env vars
144- Fix ultrathink + thinking toggle
145- Reduce unnecessary logins
146- Document --system-prompt
147- Several improvements to rendering
148- Plugins UI polish
149
150## 2.0.13
151
152- Fixed `/plugin` not working on native build
153
154## 2.0.12
155
156- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
157- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
158- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
159- `/plugin validate` command for validating plugin structure and configuration
160- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
161- Plugin documentation available at https://docs.claude.com/en/docs/claude-code/plugins
162- Comprehensive error messages and diagnostics via `/doctor` command
163- Avoid flickering in `/model` selector
164- Improvements to `/help`
165- Avoid mentioning hooks in `/resume` summaries
166- Changes to the "verbose" setting in `/config` now persist across sessions
167
168## 2.0.11
169
170- Reduced system prompt size by 1.4k tokens
171- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
172- Fixed Opus fallback rate limit errors appearing incorrectly
173- Fixed /add-dir command selecting wrong default tab
174
175## 2.0.10
176
177- Rewrote terminal renderer for buttery smooth UI
178- Enable/disable MCP servers by @mentioning, or in /mcp
179- Added tab completion for shell commands in bash mode
180- PreToolUse hooks can now modify tool inputs
181- Press Ctrl-G to edit your prompt in your system's configured text editor
182- Fixes for bash permission checks with environment variables in the command
183
184## 2.0.9
185
186- Fix regression where bash backgrounding stopped working
187
188## 2.0.8
189
190- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
191- IDE: Add drag-and-drop support for files and folders in chat
192- /context: Fix counting for thinking blocks
193- Improve message rendering for users with light themes on dark terminals
194- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
195
196## 2.0.5
197
198- IDE: Fix IME unintended message submission with Enter and Tab
199- IDE: Add "Open in Terminal" link in login screen
200- Fix unhandled OAuth expiration 401 API errors
201- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
202
203## 2.0.1
204
205- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
206- Various bug fixes and presentation improvements
207
208## 2.0.0
209
210- New native VS Code extension
211- Fresh coat of paint throughout the whole app
212- /rewind a conversation to undo code changes
213- /usage command to see plan limits
214- Tab to toggle thinking (sticky across sessions)
215- Ctrl-R to search history
216- Unshipped claude config command
217- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
218- SDK: The Claude Code SDK is now the Claude Agent SDK
219- Add subagents dynamically with `--agents` flag
220
221## 1.0.126
222
223- Enable /context command for Bedrock and Vertex
224- Add mTLS support for HTTP-based OpenTelemetry exporters
225
226## 1.0.124
227
228- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
229- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
230- No longer inform Claude of the list of allowed tools when permission is denied
231- Fixed security vulnerability in Bash tool permission checks
232- Improved VSCode extension performance for large files
233
234## 1.0.123
235
236- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
237- Fixed thinking mode triggering on negation phrases like "don't think"
238- Fixed rendering performance degradation during token streaming
239- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
240- Enhanced BashTool environment snapshot logging
241- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
242- Migrated --debug logging to a file, to enable easy tailing & filtering
243
244## 1.0.120
245
246- Fix input lag during typing, especially noticeable with large prompts
247- Improved VSCode extension command registry and sessions dialog user experience
248- Enhanced sessions dialog responsiveness and visual feedback
249- Fixed IDE compatibility issue by removing worktree support check
250- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
251
252## 1.0.119
253
254- Fix Windows issue where process visually freezes on entering interactive mode
255- Support dynamic headers for MCP servers via headersHelper configuration
256- Fix thinking mode not working in headless sessions
257- Fix slash commands now properly update allowed tools instead of replacing them
258
259## 1.0.117
260
261- Add Ctrl-R history search to recall previous commands like bash/zsh
262- Fix input lag while typing, especially on Windows
263- Add sed command to auto-allowed commands in acceptEdits mode
264- Fix Windows PATH comparison to be case-insensitive for drive letters
265- Add permissions management hint to /add-dir output
266
267## 1.0.115
268
269- Improve thinking mode display with enhanced visual effects
270- Type /t to temporarily disable thinking mode in your prompt
271- Improve path validation for glob and grep tools
272- Show condensed output for post-tool hooks to reduce visual clutter
273- Fix visual feedback when loading state completes
274- Improve UI consistency for permission request dialogs
275
276## 1.0.113
277
278- Deprecated piped input in interactive mode
279- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
280
281## 1.0.112
282
283- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
284- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
285- Hooks: Added systemMessage support for SessionEnd hooks
286- Added `spinnerTipsEnabled` setting to disable spinner tips
287- IDE: Various improvements and bug fixes
288
289## 1.0.111
290
291- /model now validates provided model names
292- Fixed Bash tool crashes caused by malformed shell syntax parsing
293
294## 1.0.110
295
296- /terminal-setup command now supports WezTerm
297- MCP: OAuth tokens now proactively refresh before expiration
298- Fixed reliability issues with background Bash processes
299
300## 1.0.109
301
302- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
303
304## 1.0.106
305
306- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
307
308## 1.0.97
309
310- Settings: /doctor now validates permission rule syntax and suggests corrections
311
312## 1.0.94
313
314- Vertex: add support for global endpoints for supported models
315- /memory command now allows direct editing of all imported memory files
316- SDK: Add custom tools as callbacks
317- Added /todos command to list current todo items
318
319## 1.0.93
320
321- Windows: Add alt + v shortcut for pasting images from clipboard
322- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
323
324## 1.0.90
325
326- Settings file changes take effect immediately - no restart required
327
328## 1.0.88
329
330- Fixed issue causing "OAuth authentication is currently not supported"
331- Status line input now includes `exceeds_200k_tokens`
332- Fixed incorrect usage tracking in /cost.
333- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
334- Bedrock: Updated default Sonnet model to Sonnet 4
335
336## 1.0.86
337
338- Added /context to help users self-serve debug context issues
339- SDK: Added UUID support for all SDK messages
340- SDK: Added `--replay-user-messages` to replay user messages back to stdout
341
342## 1.0.85
343
344- Status line input now includes session cost info
345- Hooks: Introduced SessionEnd hook
346
347## 1.0.84
348
349- Fix tool_use/tool_result id mismatch error when network is unstable
350- Fix Claude sometimes ignoring real-time steering when wrapping up a task
351- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
352- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
353
354## 1.0.83
355
356- @-mention: Support files with spaces in path
357- New shimmering spinner
358
359## 1.0.82
360
361- SDK: Add request cancellation support
362- SDK: New additionalDirectories option to search custom paths, improved slash command processing
363- Settings: Validation prevents invalid fields in .claude/settings.json files
364- MCP: Improve tool name consistency
365- Bash: Fix crash when Claude tries to automatically read large files
366
367## 1.0.81
368
369- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
370- Agents: Fix custom agent loading when agent files are unparsable
371
372## 1.0.80
373
374- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
375
376## 1.0.77
377
378- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
379- SDK: Add session support and permission denial tracking
380- Fix token limit errors in conversation summarization
381- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
382
383## 1.0.73
384
385- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
386- MCP: Press Esc to cancel OAuth authentication flows
387- Bash: Improved command validation and reduced false security warnings
388- UI: Enhanced spinner animations and status line visual hierarchy
389- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
390
391## 1.0.72
392
393- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
394
395## 1.0.71
396
397- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
398- Customizable status line: add your terminal prompt to Claude Code with /statusline
399
400## 1.0.70
401
402- Performance: Optimized message rendering for better performance with large contexts
403- Windows: Fixed native file search, ripgrep, and subagent functionality
404- Added support for @-mentions in slash command arguments
405
406## 1.0.69
407
408- Upgraded Opus to version 4.1
409
410## 1.0.68
411
412- Fix incorrect model names being used for certain commands like `/pr-comments`
413- 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.
414- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
415- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
416- SDK: Added canUseTool callback support for tool confirmation
417- Added `disableAllHooks` setting
418- Improved file suggestions performance in large repos
419
420## 1.0.65
421
422- IDE: Fixed connection stability issues and error handling for diagnostics
423- Windows: Fixed shell environment setup for users without .bashrc files
424
425## 1.0.64
426
427- Agents: Added model customization support - you can now specify which model an agent should use
428- Agents: Fixed unintended access to the recursive agent tool
429- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
430- SDK: Fixed user input tracking across multi-turn conversations
431- Added hidden files to file search and @-mention suggestions
432
433## 1.0.63
434
435- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
436
437## 1.0.62
438
439- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
440- Hooks: Added SessionStart hook for new session initialization
441- /add-dir command now supports typeahead for directory paths
442- Improved network connectivity check reliability
443
444## 1.0.61
445
446- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
447- Settings: Added `--settings` flag to load settings from a JSON file
448- Settings: Fixed resolution of settings files paths that are symlinks
449- OTEL: Fixed reporting of wrong organization after authentication changes
450- Slash commands: Fixed permissions checking for allowed-tools with Bash
451- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
452- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
453- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
454
455## 1.0.60
456
457- You can now create custom subagents for specialized tasks! Run /agents to get started
458
459## 1.0.59
460
461- SDK: Added tool confirmation support with canUseTool callback
462- SDK: Allow specifying env for spawned process
463- Hooks: Exposed PermissionDecision to hooks (including "ask")
464- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
465- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
466
467## 1.0.58
468
469- Added support for reading PDFs
470- MCP: Improved server health status display in 'claude mcp list'
471- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
472
473## 1.0.57
474
475- Added support for specifying a model in slash commands
476- Improved permission messages to help Claude understand allowed tools
477- Fix: Remove trailing newlines from bash output in terminal wrapping
478
479## 1.0.56
480
481- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
482- Fixes for WSL IDE detection
483- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
484
485## 1.0.55
486
487- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
488- Windows: fixed Ctrl+Z crash
489- SDK: Added ability to capture error logging
490- Add --system-prompt-file option to override system prompt in print mode
491
492## 1.0.54
493
494- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
495- Custom slash commands: Added argument-hint to frontmatter
496- Windows: OAuth uses port 45454 and properly constructs browser URL
497- Windows: mode switching now uses alt + m, and plan mode renders properly
498- Shell: Switch to in-memory shell snapshot to fix file-related errors
499
500## 1.0.53
501
502- Updated @-mention file truncation from 100 lines to 2000 lines
503- 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).
504
505## 1.0.52
506
507- Added support for MCP server instructions
508
509## 1.0.51
510
511- Added support for native Windows (requires Git for Windows)
512- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
513- Settings: /doctor can now help you identify and fix invalid setting files
514- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
515- Increased auto-compact warning threshold from 60% to 80%
516- Fixed an issue with handling user directories with spaces for shell snapshots
517- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
518- Custom slash commands: Fixed user-level commands in subdirectories
519- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
520
521## 1.0.48
522
523- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
524- Added progress messages to Bash tool based on the last 5 lines of command output
525- Added expanding variables support for MCP server configuration
526- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
527- Improved IDE extension path handling when Claude Code runs in WSL
528- Hooks: Added a PreCompact hook
529- Vim mode: Added c, f/F, t/T
530
531## 1.0.45
532
533- Redesigned Search (Grep) tool with new tool input parameters and features
534- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
535- Fixed config file corruption issue by enforcing atomic writes
536- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
537- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
538- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
539
540## 1.0.44
541
542- New /export command lets you quickly export a conversation for sharing
543- MCP: resource_link tool results are now supported
544- MCP: tool annotations and tool titles now display in /mcp view
545- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
546
547## 1.0.43
548
549- Fixed a bug where the theme selector was saving excessively
550- Hooks: Added EPIPE system error handling
551
552## 1.0.42
553
554- Added tilde (`~`) expansion support to `/add-dir` command
555
556## 1.0.41
557
558- Hooks: Split Stop hook triggering into Stop and SubagentStop
559- Hooks: Enabled optional timeout configuration for each command
560- Hooks: Added "hook_event_name" to hook input
561- Fixed a bug where MCP tools would display twice in tool list
562- New tool parameters JSON for Bash tool in `tool_decision` event
563
564## 1.0.40
565
566- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
567
568## 1.0.39
569
570- New Active Time metric in OpenTelemetry logging
571
572## 1.0.38
573
574- 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
575
576## 1.0.37
577
578- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
579
580## 1.0.36
581
582- Web search now takes today's date into context
583- Fixed a bug where stdio MCP servers were not terminating properly on exit
584
585## 1.0.35
586
587- Added support for MCP OAuth Authorization Server discovery
588
589## 1.0.34
590
591- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
592
593## 1.0.33
594
595- Improved logging functionality with session ID support
596- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
597- Improvements to plan mode
598
599## 1.0.32
600
601- Updated loopback config for litellm
602- Added forceLoginMethod setting to bypass login selection screen
603
604## 1.0.31
605
606- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
607
608## 1.0.30
609
610- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
611- Improved file path autocomplete with filename matching
612- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
613- Enhanced jq regex support for complex filters with pipes and select
614
615## 1.0.29
616
617- Improved CJK character support in cursor navigation and rendering
618
619## 1.0.28
620
621- Slash commands: Fix selector display during history navigation
622- Resizes images before upload to prevent API size limit errors
623- Added XDG_CONFIG_HOME support to configuration directory
624- Performance optimizations for memory usage
625- New attributes (terminal.type, language) in OpenTelemetry logging
626
627## 1.0.27
628
629- Streamable HTTP MCP servers are now supported
630- Remote MCP servers (SSE and HTTP) now support OAuth
631- MCP resources can now be @-mentioned
632- /resume slash command to switch conversations within Claude Code
633
634## 1.0.25
635
636- Slash commands: moved "project" and "user" prefixes to descriptions
637- Slash commands: improved reliability for command discovery
638- Improved support for Ghostty
639- Improved web search reliability
640
641## 1.0.24
642
643- Improved /mcp output
644- Fixed a bug where settings arrays got overwritten instead of merged
645
646## 1.0.23
647
648- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
649- Released Python SDK: pip install claude-code-sdk to get started
650
651## 1.0.22
652
653- SDK: Renamed `total_cost` to `total_cost_usd`
654
655## 1.0.21
656
657- Improved editing of files with tab-based indentation
658- Fix for tool_use without matching tool_result errors
659- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
660
661## 1.0.18
662
663- Added --add-dir CLI argument for specifying additional working directories
664- Added streaming input support without require -p flag
665- Improved startup performance and session storage performance
666- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
667- Added detailed MCP server tools display (/mcp)
668- MCP authentication and permission improvements
669- Added auto-reconnection for MCP SSE connections on disconnect
670- Fixed issue where pasted content was lost when dialogs appeared
671
672## 1.0.17
673
674- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
675- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
676- MCP server list UI improvements
677- Update Claude Code process title to display "claude" instead of "node"
678
679## 1.0.11
680
681- Claude Code can now also be used with a Claude Pro subscription
682- Added /upgrade for smoother switching to Claude Max plans
683- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
684- Improved shell configuration error handling
685- Improved todo list handling during compaction
686
687## 1.0.10
688
689- Added markdown table support
690- Improved streaming performance
691
692## 1.0.8
693
694- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
695- Increased default otel interval from 1s -> 5s
696- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
697- Fixed a regression where search tools unnecessarily asked for permissions
698- Added support for triggering thinking non-English languages
699- Improved compacting UI
700
701## 1.0.7
702
703- Renamed /allowed-tools -> /permissions
704- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
705- Deprecated claude config commands in favor of editing settings.json
706- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
707- Improved error handling for /install-github-app
708- Bugfixes, UI polish, and tool reliability improvements
709
710## 1.0.6
711
712- Improved edit reliability for tab-indented files
713- Respect CLAUDE_CONFIG_DIR everywhere
714- Reduced unnecessary tool permission prompts
715- Added support for symlinks in @file typeahead
716- Bugfixes, UI polish, and tool reliability improvements
717
718## 1.0.4
719
720- Fixed a bug where MCP tool errors weren't being parsed correctly
721
722## 1.0.1
723
724- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
725- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
726- Updated documentation links and OAuth process descriptions
727
728## 1.0.0
729
730- Claude Code is now generally available
731- Introducing Sonnet 4 and Opus 4 models
732
733## 0.2.125
734
735- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
736- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
737
738## 0.2.117
739
740- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
741- Introduced settings.cleanupPeriodDays
742- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
743- Introduced --debug mode
744
745## 0.2.108
746
747- You can now send messages to Claude while it works to steer Claude in real-time
748- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
749- Fixed a bug where thinking was not working in -p mode
750- Fixed a regression in /cost reporting
751- Deprecated MCP wizard interface in favor of other MCP commands
752- Lots of other bugfixes and improvements
753
754## 0.2.107
755
756- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
757
758## 0.2.106
759
760- MCP SSE server configs can now specify custom headers
761- Fixed a bug where MCP permission prompt didn't always show correctly
762
763## 0.2.105
764
765- Claude can now search the web
766- Moved system & account status to /status
767- Added word movement keybindings for Vim
768- Improved latency for startup, todo tool, and file edits
769
770## 0.2.102
771
772- Improved thinking triggering reliability
773- Improved @mention reliability for images and folders
774- You can now paste multiple large chunks into one prompt
775
776## 0.2.100
777
778- Fixed a crash caused by a stack overflow error
779- Made db storage optional; missing db support disables --continue and --resume
780
781## 0.2.98
782
783- Fixed an issue where auto-compact was running twice
784
785## 0.2.96
786
787- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
788
789## 0.2.93
790
791- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
792- Claude now has access to a Todo list that helps it stay on track and be more organized
793
794## 0.2.82
795
796- Added support for --disallowedTools
797- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
798
799## 0.2.75
800
801- Hit Enter to queue up additional messages while Claude is working
802- Drag in or copy/paste image files directly into the prompt
803- @-mention files to directly add them to context
804- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
805- Improved performance for filename auto-complete
806
807## 0.2.74
808
809- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
810- Task tool can now perform writes and run bash commands
811
812## 0.2.72
813
814- Updated spinner to indicate tokens loaded and tool usage
815
816## 0.2.70
817
818- Network commands like curl are now available for Claude to use
819- Claude can now run multiple web queries in parallel
820- Pressing ESC once immediately interrupts Claude in Auto-accept mode
821
822## 0.2.69
823
824- Fixed UI glitches with improved Select component behavior
825- Enhanced terminal output display with better text truncation logic
826
827## 0.2.67
828
829- Shared project permission rules can be saved in .claude/settings.json
830
831## 0.2.66
832
833- Print mode (-p) now supports streaming output via --output-format=stream-json
834- Fixed issue where pasting could trigger memory or bash mode unexpectedly
835
836## 0.2.63
837
838- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
839
840## 0.2.61
841
842- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
843- Enhanced image detection for more reliable clipboard paste functionality
844- Fixed an issue where ESC key could crash the conversation history selector
845
846## 0.2.59
847
848- Copy+paste images directly into your prompt
849- Improved progress indicators for bash and fetch tools
850- Bugfixes for non-interactive mode (-p)
851
852## 0.2.54
853
854- Quickly add to Memory by starting your message with '#'
855- Press ctrl+r to see full output for long tool results
856- Added support for MCP SSE transport
857
858## 0.2.53
859
860- New web fetch tool lets Claude view URLs that you paste in
861- Fixed a bug with JPEG detection
862
863## 0.2.50
864
865- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
866
867## 0.2.49
868
869- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
870
871## 0.2.47
872
873- Press Tab to auto-complete file and folder names
874- Press Shift + Tab to toggle auto-accept for file edits
875- Automatic conversation compaction for infinite conversation length (toggle with /config)
876
877## 0.2.44
878
879- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
880
881## 0.2.41
882
883- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
884- MCP server startup no longer blocks the app from starting up
885
886## 0.2.37
887
888- New /release-notes command lets you view release notes at any time
889- `claude config add/remove` commands now accept multiple values separated by commas or spaces
890
891## 0.2.36
892
893- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
894- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
895
896## 0.2.34
897
898- Vim bindings for text input - enable with /vim or /config
899
900## 0.2.32
901
902- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
903- Fix for some PersistentShell issues
904
905## 0.2.31
906
907- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
908- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
909
910## 0.2.30
911
912- Added ANSI color theme for better terminal compatibility
913- Fixed issue where slash command arguments weren't being sent properly
914- (Mac-only) API keys are now stored in macOS Keychain
915
916## 0.2.26
917
918- New /approved-tools command for managing tool permissions
919- Word-level diff display for improved code readability
920- Fuzzy matching for slash commands
921
922## 0.2.21
923
924- Fuzzy matching for /commands