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 0ea180e · on Oct 17, 2025 History
823 lines · 29.6 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.22
4
5- Edit the plan in your text editor using CTRL+G
6- Fixed content layout shift when scrolling through slash commands
7- IDE: Add toggle to enable/disable thinking.
8- Fix bug causing duplicate permission prompts with parallel tool calls
9- Add support for enterprise managed MCP allowlist and denylist
10- Support MCP `structuredContent` field in tool responses
11- Added an interactive question tool
12- Claude will now ask you questions more often in plan mode
13- Added Haiku 4.5 as a model option for Pro users
14- Fixed an issue where queued commands don't have access to previous messages' output
15
16## 2.0.20
17
18- Added support for Claude Skills
19
20## 2.0.19
21
22- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
23- Fixed a bug where Haiku was unnecessarily called in print mode
24
25## 2.0.17
26
27- Added Haiku 4.5 to model selector!
28- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
29- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
30- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
31- OTEL: support HTTP_PROXY and HTTPS_PROXY
32- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
33
34## 2.0.15
35
36- Fixed bug with resuming where previously created files needed to be read again before writing
37- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
38
39## 2.0.14
40
41- Fix @-mentioning MCP servers to toggle them on/off
42- Improve permission checks for bash with inline env vars
43- Fix ultrathink + thinking toggle
44- Reduce unnecessary logins
45- Document --system-prompt
46- Several improvements to rendering
47- Plugins UI polish
48
49## 2.0.13
50
51- Fixed `/plugin` not working on native build
52
53## 2.0.12
54
55- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
56- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
57- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
58- `/plugin validate` command for validating plugin structure and configuration
59- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
60- Plugin documentation available at https://docs.claude.com/en/docs/claude-code/plugins
61- Comprehensive error messages and diagnostics via `/doctor` command
62- Avoid flickering in `/model` selector
63- Improvements to `/help`
64- Avoid mentioning hooks in `/resume` summaries
65- Changes to the "verbose" setting in `/config` now persist across sessions
66
67## 2.0.11
68
69- Reduced system prompt size by 1.4k tokens
70- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
71- Fixed Opus fallback rate limit errors appearing incorrectly
72- Fixed /add-dir command selecting wrong default tab
73
74## 2.0.10
75
76- Rewrote terminal renderer for buttery smooth UI
77- Enable/disable MCP servers by @mentioning, or in /mcp
78- Added tab completion for shell commands in bash mode
79- PreToolUse hooks can now modify tool inputs
80- Press Ctrl-G to edit your prompt in your system's configured text editor
81- Fixes for bash permission checks with environment variables in the command
82
83## 2.0.9
84
85- Fix regression where bash backgrounding stopped working
86
87## 2.0.8
88
89- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
90- IDE: Add drag-and-drop support for files and folders in chat
91- /context: Fix counting for thinking blocks
92- Improve message rendering for users with light themes on dark terminals
93- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
94
95## 2.0.5
96
97- IDE: Fix IME unintended message submission with Enter and Tab
98- IDE: Add "Open in Terminal" link in login screen
99- Fix unhandled OAuth expiration 401 API errors
100- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
101
102## 2.0.1
103
104- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
105- Various bug fixes and presentation improvements
106
107## 2.0.0
108
109- New native VS Code extension
110- Fresh coat of paint throughout the whole app
111- /rewind a conversation to undo code changes
112- /usage command to see plan limits
113- Tab to toggle thinking (sticky across sessions)
114- Ctrl-R to search history
115- Unshipped claude config command
116- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
117- SDK: The Claude Code SDK is now the Claude Agent SDK
118- Add subagents dynamically with `--agents` flag
119
120## 1.0.126
121
122- Enable /context command for Bedrock and Vertex
123- Add mTLS support for HTTP-based OpenTelemetry exporters
124
125## 1.0.124
126
127- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
128- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
129- No longer inform Claude of the list of allowed tools when permission is denied
130- Fixed security vulnerability in Bash tool permission checks
131- Improved VSCode extension performance for large files
132
133## 1.0.123
134
135- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
136- Fixed thinking mode triggering on negation phrases like "don't think"
137- Fixed rendering performance degradation during token streaming
138- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
139- Enhanced BashTool environment snapshot logging
140- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
141- Migrated --debug logging to a file, to enable easy tailing & filtering
142
143## 1.0.120
144
145- Fix input lag during typing, especially noticeable with large prompts
146- Improved VSCode extension command registry and sessions dialog user experience
147- Enhanced sessions dialog responsiveness and visual feedback
148- Fixed IDE compatibility issue by removing worktree support check
149- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
150
151## 1.0.119
152
153- Fix Windows issue where process visually freezes on entering interactive mode
154- Support dynamic headers for MCP servers via headersHelper configuration
155- Fix thinking mode not working in headless sessions
156- Fix slash commands now properly update allowed tools instead of replacing them
157
158## 1.0.117
159
160- Add Ctrl-R history search to recall previous commands like bash/zsh
161- Fix input lag while typing, especially on Windows
162- Add sed command to auto-allowed commands in acceptEdits mode
163- Fix Windows PATH comparison to be case-insensitive for drive letters
164- Add permissions management hint to /add-dir output
165
166## 1.0.115
167
168- Improve thinking mode display with enhanced visual effects
169- Type /t to temporarily disable thinking mode in your prompt
170- Improve path validation for glob and grep tools
171- Show condensed output for post-tool hooks to reduce visual clutter
172- Fix visual feedback when loading state completes
173- Improve UI consistency for permission request dialogs
174
175## 1.0.113
176
177- Deprecated piped input in interactive mode
178- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
179
180## 1.0.112
181
182- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
183- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
184- Hooks: Added systemMessage support for SessionEnd hooks
185- Added `spinnerTipsEnabled` setting to disable spinner tips
186- IDE: Various improvements and bug fixes
187
188## 1.0.111
189
190- /model now validates provided model names
191- Fixed Bash tool crashes caused by malformed shell syntax parsing
192
193## 1.0.110
194
195- /terminal-setup command now supports WezTerm
196- MCP: OAuth tokens now proactively refresh before expiration
197- Fixed reliability issues with background Bash processes
198
199## 1.0.109
200
201- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
202
203## 1.0.106
204
205- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
206
207## 1.0.97
208
209- Settings: /doctor now validates permission rule syntax and suggests corrections
210
211## 1.0.94
212
213- Vertex: add support for global endpoints for supported models
214- /memory command now allows direct editing of all imported memory files
215- SDK: Add custom tools as callbacks
216- Added /todos command to list current todo items
217
218## 1.0.93
219
220- Windows: Add alt + v shortcut for pasting images from clipboard
221- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
222
223## 1.0.90
224
225- Settings file changes take effect immediately - no restart required
226
227## 1.0.88
228
229- Fixed issue causing "OAuth authentication is currently not supported"
230- Status line input now includes `exceeds_200k_tokens`
231- Fixed incorrect usage tracking in /cost.
232- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
233- Bedrock: Updated default Sonnet model to Sonnet 4
234
235## 1.0.86
236
237- Added /context to help users self-serve debug context issues
238- SDK: Added UUID support for all SDK messages
239- SDK: Added `--replay-user-messages` to replay user messages back to stdout
240
241## 1.0.85
242
243- Status line input now includes session cost info
244- Hooks: Introduced SessionEnd hook
245
246## 1.0.84
247
248- Fix tool_use/tool_result id mismatch error when network is unstable
249- Fix Claude sometimes ignoring real-time steering when wrapping up a task
250- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
251- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
252
253## 1.0.83
254
255- @-mention: Support files with spaces in path
256- New shimmering spinner
257
258## 1.0.82
259
260- SDK: Add request cancellation support
261- SDK: New additionalDirectories option to search custom paths, improved slash command processing
262- Settings: Validation prevents invalid fields in .claude/settings.json files
263- MCP: Improve tool name consistency
264- Bash: Fix crash when Claude tries to automatically read large files
265
266## 1.0.81
267
268- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
269- Agents: Fix custom agent loading when agent files are unparsable
270
271## 1.0.80
272
273- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
274
275## 1.0.77
276
277- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
278- SDK: Add session support and permission denial tracking
279- Fix token limit errors in conversation summarization
280- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
281
282## 1.0.73
283
284- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
285- MCP: Press Esc to cancel OAuth authentication flows
286- Bash: Improved command validation and reduced false security warnings
287- UI: Enhanced spinner animations and status line visual hierarchy
288- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
289
290## 1.0.72
291
292- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
293
294## 1.0.71
295
296- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
297- Customizable status line: add your terminal prompt to Claude Code with /statusline
298
299## 1.0.70
300
301- Performance: Optimized message rendering for better performance with large contexts
302- Windows: Fixed native file search, ripgrep, and subagent functionality
303- Added support for @-mentions in slash command arguments
304
305## 1.0.69
306
307- Upgraded Opus to version 4.1
308
309## 1.0.68
310
311- Fix incorrect model names being used for certain commands like `/pr-comments`
312- 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.
313- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
314- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
315- SDK: Added canUseTool callback support for tool confirmation
316- Added `disableAllHooks` setting
317- Improved file suggestions performance in large repos
318
319## 1.0.65
320
321- IDE: Fixed connection stability issues and error handling for diagnostics
322- Windows: Fixed shell environment setup for users without .bashrc files
323
324## 1.0.64
325
326- Agents: Added model customization support - you can now specify which model an agent should use
327- Agents: Fixed unintended access to the recursive agent tool
328- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
329- SDK: Fixed user input tracking across multi-turn conversations
330- Added hidden files to file search and @-mention suggestions
331
332## 1.0.63
333
334- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
335
336## 1.0.62
337
338- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
339- Hooks: Added SessionStart hook for new session initialization
340- /add-dir command now supports typeahead for directory paths
341- Improved network connectivity check reliability
342
343## 1.0.61
344
345- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
346- Settings: Added `--settings` flag to load settings from a JSON file
347- Settings: Fixed resolution of settings files paths that are symlinks
348- OTEL: Fixed reporting of wrong organization after authentication changes
349- Slash commands: Fixed permissions checking for allowed-tools with Bash
350- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
351- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
352- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
353
354## 1.0.60
355
356- You can now create custom subagents for specialized tasks! Run /agents to get started
357
358## 1.0.59
359
360- SDK: Added tool confirmation support with canUseTool callback
361- SDK: Allow specifying env for spawned process
362- Hooks: Exposed PermissionDecision to hooks (including "ask")
363- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
364- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
365
366## 1.0.58
367
368- Added support for reading PDFs
369- MCP: Improved server health status display in 'claude mcp list'
370- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
371
372## 1.0.57
373
374- Added support for specifying a model in slash commands
375- Improved permission messages to help Claude understand allowed tools
376- Fix: Remove trailing newlines from bash output in terminal wrapping
377
378## 1.0.56
379
380- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
381- Fixes for WSL IDE detection
382- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
383
384## 1.0.55
385
386- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
387- Windows: fixed Ctrl+Z crash
388- SDK: Added ability to capture error logging
389- Add --system-prompt-file option to override system prompt in print mode
390
391## 1.0.54
392
393- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
394- Custom slash commands: Added argument-hint to frontmatter
395- Windows: OAuth uses port 45454 and properly constructs browser URL
396- Windows: mode switching now uses alt + m, and plan mode renders properly
397- Shell: Switch to in-memory shell snapshot to fix file-related errors
398
399## 1.0.53
400
401- Updated @-mention file truncation from 100 lines to 2000 lines
402- 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).
403
404## 1.0.52
405
406- Added support for MCP server instructions
407
408## 1.0.51
409
410- Added support for native Windows (requires Git for Windows)
411- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
412- Settings: /doctor can now help you identify and fix invalid setting files
413- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
414- Increased auto-compact warning threshold from 60% to 80%
415- Fixed an issue with handling user directories with spaces for shell snapshots
416- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
417- Custom slash commands: Fixed user-level commands in subdirectories
418- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
419
420## 1.0.48
421
422- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
423- Added progress messages to Bash tool based on the last 5 lines of command output
424- Added expanding variables support for MCP server configuration
425- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
426- Improved IDE extension path handling when Claude Code runs in WSL
427- Hooks: Added a PreCompact hook
428- Vim mode: Added c, f/F, t/T
429
430## 1.0.45
431
432- Redesigned Search (Grep) tool with new tool input parameters and features
433- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
434- Fixed config file corruption issue by enforcing atomic writes
435- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
436- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
437- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
438
439## 1.0.44
440
441- New /export command lets you quickly export a conversation for sharing
442- MCP: resource_link tool results are now supported
443- MCP: tool annotations and tool titles now display in /mcp view
444- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
445
446## 1.0.43
447
448- Fixed a bug where the theme selector was saving excessively
449- Hooks: Added EPIPE system error handling
450
451## 1.0.42
452
453- Added tilde (`~`) expansion support to `/add-dir` command
454
455## 1.0.41
456
457- Hooks: Split Stop hook triggering into Stop and SubagentStop
458- Hooks: Enabled optional timeout configuration for each command
459- Hooks: Added "hook_event_name" to hook input
460- Fixed a bug where MCP tools would display twice in tool list
461- New tool parameters JSON for Bash tool in `tool_decision` event
462
463## 1.0.40
464
465- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
466
467## 1.0.39
468
469- New Active Time metric in OpenTelemetry logging
470
471## 1.0.38
472
473- 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
474
475## 1.0.37
476
477- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
478
479## 1.0.36
480
481- Web search now takes today's date into context
482- Fixed a bug where stdio MCP servers were not terminating properly on exit
483
484## 1.0.35
485
486- Added support for MCP OAuth Authorization Server discovery
487
488## 1.0.34
489
490- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
491
492## 1.0.33
493
494- Improved logging functionality with session ID support
495- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
496- Improvements to plan mode
497
498## 1.0.32
499
500- Updated loopback config for litellm
501- Added forceLoginMethod setting to bypass login selection screen
502
503## 1.0.31
504
505- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
506
507## 1.0.30
508
509- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
510- Improved file path autocomplete with filename matching
511- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
512- Enhanced jq regex support for complex filters with pipes and select
513
514## 1.0.29
515
516- Improved CJK character support in cursor navigation and rendering
517
518## 1.0.28
519
520- Slash commands: Fix selector display during history navigation
521- Resizes images before upload to prevent API size limit errors
522- Added XDG_CONFIG_HOME support to configuration directory
523- Performance optimizations for memory usage
524- New attributes (terminal.type, language) in OpenTelemetry logging
525
526## 1.0.27
527
528- Streamable HTTP MCP servers are now supported
529- Remote MCP servers (SSE and HTTP) now support OAuth
530- MCP resources can now be @-mentioned
531- /resume slash command to switch conversations within Claude Code
532
533## 1.0.25
534
535- Slash commands: moved "project" and "user" prefixes to descriptions
536- Slash commands: improved reliability for command discovery
537- Improved support for Ghostty
538- Improved web search reliability
539
540## 1.0.24
541
542- Improved /mcp output
543- Fixed a bug where settings arrays got overwritten instead of merged
544
545## 1.0.23
546
547- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
548- Released Python SDK: pip install claude-code-sdk to get started
549
550## 1.0.22
551
552- SDK: Renamed `total_cost` to `total_cost_usd`
553
554## 1.0.21
555
556- Improved editing of files with tab-based indentation
557- Fix for tool_use without matching tool_result errors
558- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
559
560## 1.0.18
561
562- Added --add-dir CLI argument for specifying additional working directories
563- Added streaming input support without require -p flag
564- Improved startup performance and session storage performance
565- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
566- Added detailed MCP server tools display (/mcp)
567- MCP authentication and permission improvements
568- Added auto-reconnection for MCP SSE connections on disconnect
569- Fixed issue where pasted content was lost when dialogs appeared
570
571## 1.0.17
572
573- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
574- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
575- MCP server list UI improvements
576- Update Claude Code process title to display "claude" instead of "node"
577
578## 1.0.11
579
580- Claude Code can now also be used with a Claude Pro subscription
581- Added /upgrade for smoother switching to Claude Max plans
582- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
583- Improved shell configuration error handling
584- Improved todo list handling during compaction
585
586## 1.0.10
587
588- Added markdown table support
589- Improved streaming performance
590
591## 1.0.8
592
593- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
594- Increased default otel interval from 1s -> 5s
595- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
596- Fixed a regression where search tools unnecessarily asked for permissions
597- Added support for triggering thinking non-English languages
598- Improved compacting UI
599
600## 1.0.7
601
602- Renamed /allowed-tools -> /permissions
603- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
604- Deprecated claude config commands in favor of editing settings.json
605- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
606- Improved error handling for /install-github-app
607- Bugfixes, UI polish, and tool reliability improvements
608
609## 1.0.6
610
611- Improved edit reliability for tab-indented files
612- Respect CLAUDE_CONFIG_DIR everywhere
613- Reduced unnecessary tool permission prompts
614- Added support for symlinks in @file typeahead
615- Bugfixes, UI polish, and tool reliability improvements
616
617## 1.0.4
618
619- Fixed a bug where MCP tool errors weren't being parsed correctly
620
621## 1.0.1
622
623- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
624- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
625- Updated documentation links and OAuth process descriptions
626
627## 1.0.0
628
629- Claude Code is now generally available
630- Introducing Sonnet 4 and Opus 4 models
631
632## 0.2.125
633
634- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
635- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
636
637## 0.2.117
638
639- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
640- Introduced settings.cleanupPeriodDays
641- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
642- Introduced --debug mode
643
644## 0.2.108
645
646- You can now send messages to Claude while it works to steer Claude in real-time
647- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
648- Fixed a bug where thinking was not working in -p mode
649- Fixed a regression in /cost reporting
650- Deprecated MCP wizard interface in favor of other MCP commands
651- Lots of other bugfixes and improvements
652
653## 0.2.107
654
655- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
656
657## 0.2.106
658
659- MCP SSE server configs can now specify custom headers
660- Fixed a bug where MCP permission prompt didn't always show correctly
661
662## 0.2.105
663
664- Claude can now search the web
665- Moved system & account status to /status
666- Added word movement keybindings for Vim
667- Improved latency for startup, todo tool, and file edits
668
669## 0.2.102
670
671- Improved thinking triggering reliability
672- Improved @mention reliability for images and folders
673- You can now paste multiple large chunks into one prompt
674
675## 0.2.100
676
677- Fixed a crash caused by a stack overflow error
678- Made db storage optional; missing db support disables --continue and --resume
679
680## 0.2.98
681
682- Fixed an issue where auto-compact was running twice
683
684## 0.2.96
685
686- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
687
688## 0.2.93
689
690- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
691- Claude now has access to a Todo list that helps it stay on track and be more organized
692
693## 0.2.82
694
695- Added support for --disallowedTools
696- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
697
698## 0.2.75
699
700- Hit Enter to queue up additional messages while Claude is working
701- Drag in or copy/paste image files directly into the prompt
702- @-mention files to directly add them to context
703- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
704- Improved performance for filename auto-complete
705
706## 0.2.74
707
708- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
709- Task tool can now perform writes and run bash commands
710
711## 0.2.72
712
713- Updated spinner to indicate tokens loaded and tool usage
714
715## 0.2.70
716
717- Network commands like curl are now available for Claude to use
718- Claude can now run multiple web queries in parallel
719- Pressing ESC once immediately interrupts Claude in Auto-accept mode
720
721## 0.2.69
722
723- Fixed UI glitches with improved Select component behavior
724- Enhanced terminal output display with better text truncation logic
725
726## 0.2.67
727
728- Shared project permission rules can be saved in .claude/settings.json
729
730## 0.2.66
731
732- Print mode (-p) now supports streaming output via --output-format=stream-json
733- Fixed issue where pasting could trigger memory or bash mode unexpectedly
734
735## 0.2.63
736
737- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
738
739## 0.2.61
740
741- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
742- Enhanced image detection for more reliable clipboard paste functionality
743- Fixed an issue where ESC key could crash the conversation history selector
744
745## 0.2.59
746
747- Copy+paste images directly into your prompt
748- Improved progress indicators for bash and fetch tools
749- Bugfixes for non-interactive mode (-p)
750
751## 0.2.54
752
753- Quickly add to Memory by starting your message with '#'
754- Press ctrl+r to see full output for long tool results
755- Added support for MCP SSE transport
756
757## 0.2.53
758
759- New web fetch tool lets Claude view URLs that you paste in
760- Fixed a bug with JPEG detection
761
762## 0.2.50
763
764- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
765
766## 0.2.49
767
768- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
769
770## 0.2.47
771
772- Press Tab to auto-complete file and folder names
773- Press Shift + Tab to toggle auto-accept for file edits
774- Automatic conversation compaction for infinite conversation length (toggle with /config)
775
776## 0.2.44
777
778- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
779
780## 0.2.41
781
782- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
783- MCP server startup no longer blocks the app from starting up
784
785## 0.2.37
786
787- New /release-notes command lets you view release notes at any time
788- `claude config add/remove` commands now accept multiple values separated by commas or spaces
789
790## 0.2.36
791
792- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
793- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
794
795## 0.2.34
796
797- Vim bindings for text input - enable with /vim or /config
798
799## 0.2.32
800
801- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
802- Fix for some PersistentShell issues
803
804## 0.2.31
805
806- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
807- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
808
809## 0.2.30
810
811- Added ANSI color theme for better terminal compatibility
812- Fixed issue where slash command arguments weren't being sent properly
813- (Mac-only) API keys are now stored in macOS Keychain
814
815## 0.2.26
816
817- New /approved-tools command for managing tool permissions
818- Word-level diff display for improved code readability
819- Fuzzy matching for slash commands
820
821## 0.2.21
822
823- Fuzzy matching for /commands