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 c1cd236 · on Oct 8, 2025 History
759 lines · 27.0 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.11
4
5- Reduced system prompt size by 1.4k tokens
6- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
7- Fixed Opus fallback rate limit errors appearing incorrectly
8- Fixed /add-dir command selecting wrong default tab
9
10## 2.0.10
11
12- Rewrote terminal renderer for buttery smooth UI
13- Enable/disable MCP servers by @mentioning, or in /mcp
14- Added tab completion for shell commands in bash mode
15- PreToolUse hooks can now modify tool inputs
16- Press Ctrl-G to edit your prompt in your system's configured text editor
17- Fixes for bash permission checks with environment variables in the command
18
19## 2.0.9
20
21- Fix regression where bash backgrounding stopped working
22
23## 2.0.8
24
25- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
26- IDE: Add drag-and-drop support for files and folders in chat
27- /context: Fix counting for thinking blocks
28- Improve message rendering for users with light themes on dark terminals
29- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
30
31## 2.0.5
32
33- IDE: Fix IME unintended message submission with Enter and Tab
34- IDE: Add "Open in Terminal" link in login screen
35- Fix unhandled OAuth expiration 401 API errors
36- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
37
38## 2.0.1
39
40- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
41- Various bug fixes and presentation improvements
42
43## 2.0.0
44
45- New native VS Code extension
46- Fresh coat of paint throughout the whole app
47- /rewind a conversation to undo code changes
48- /usage command to see plan limits
49- Tab to toggle thinking (sticky across sessions)
50- Ctrl-R to search history
51- Unshipped claude config command
52- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
53- SDK: The Claude Code SDK is now the Claude Agent SDK
54- Add subagents dynamically with `--agents` flag
55
56## 1.0.126
57
58- Enable /context command for Bedrock and Vertex
59- Add mTLS support for HTTP-based OpenTelemetry exporters
60
61## 1.0.124
62
63- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
64- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
65- No longer inform Claude of the list of allowed tools when permission is denied
66- Fixed security vulnerability in Bash tool permission checks
67- Improved VSCode extension performance for large files
68
69## 1.0.123
70
71- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
72- Fixed thinking mode triggering on negation phrases like "don't think"
73- Fixed rendering performance degradation during token streaming
74- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
75- Enhanced BashTool environment snapshot logging
76- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
77- Migrated --debug logging to a file, to enable easy tailing & filtering
78
79## 1.0.120
80
81- Fix input lag during typing, especially noticeable with large prompts
82- Improved VSCode extension command registry and sessions dialog user experience
83- Enhanced sessions dialog responsiveness and visual feedback
84- Fixed IDE compatibility issue by removing worktree support check
85- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
86
87## 1.0.119
88
89- Fix Windows issue where process visually freezes on entering interactive mode
90- Support dynamic headers for MCP servers via headersHelper configuration
91- Fix thinking mode not working in headless sessions
92- Fix slash commands now properly update allowed tools instead of replacing them
93
94## 1.0.117
95
96- Add Ctrl-R history search to recall previous commands like bash/zsh
97- Fix input lag while typing, especially on Windows
98- Add sed command to auto-allowed commands in acceptEdits mode
99- Fix Windows PATH comparison to be case-insensitive for drive letters
100- Add permissions management hint to /add-dir output
101
102## 1.0.115
103
104- Improve thinking mode display with enhanced visual effects
105- Type /t to temporarily disable thinking mode in your prompt
106- Improve path validation for glob and grep tools
107- Show condensed output for post-tool hooks to reduce visual clutter
108- Fix visual feedback when loading state completes
109- Improve UI consistency for permission request dialogs
110
111## 1.0.113
112
113- Deprecated piped input in interactive mode
114- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
115
116## 1.0.112
117
118- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
119- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
120- Hooks: Added systemMessage support for SessionEnd hooks
121- Added `spinnerTipsEnabled` setting to disable spinner tips
122- IDE: Various improvements and bug fixes
123
124## 1.0.111
125
126- /model now validates provided model names
127- Fixed Bash tool crashes caused by malformed shell syntax parsing
128
129## 1.0.110
130
131- /terminal-setup command now supports WezTerm
132- MCP: OAuth tokens now proactively refresh before expiration
133- Fixed reliability issues with background Bash processes
134
135## 1.0.109
136
137- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
138
139## 1.0.106
140
141- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
142
143## 1.0.97
144
145- Settings: /doctor now validates permission rule syntax and suggests corrections
146
147## 1.0.94
148
149- Vertex: add support for global endpoints for supported models
150- /memory command now allows direct editing of all imported memory files
151- SDK: Add custom tools as callbacks
152- Added /todos command to list current todo items
153
154## 1.0.93
155
156- Windows: Add alt + v shortcut for pasting images from clipboard
157- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
158
159## 1.0.90
160
161- Settings file changes take effect immediately - no restart required
162
163## 1.0.88
164
165- Fixed issue causing "OAuth authentication is currently not supported"
166- Status line input now includes `exceeds_200k_tokens`
167- Fixed incorrect usage tracking in /cost.
168- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
169- Bedrock: Updated default Sonnet model to Sonnet 4
170
171## 1.0.86
172
173- Added /context to help users self-serve debug context issues
174- SDK: Added UUID support for all SDK messages
175- SDK: Added `--replay-user-messages` to replay user messages back to stdout
176
177## 1.0.85
178
179- Status line input now includes session cost info
180- Hooks: Introduced SessionEnd hook
181
182## 1.0.84
183
184- Fix tool_use/tool_result id mismatch error when network is unstable
185- Fix Claude sometimes ignoring real-time steering when wrapping up a task
186- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
187- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
188
189## 1.0.83
190
191- @-mention: Support files with spaces in path
192- New shimmering spinner
193
194## 1.0.82
195
196- SDK: Add request cancellation support
197- SDK: New additionalDirectories option to search custom paths, improved slash command processing
198- Settings: Validation prevents invalid fields in .claude/settings.json files
199- MCP: Improve tool name consistency
200- Bash: Fix crash when Claude tries to automatically read large files
201
202## 1.0.81
203
204- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
205- Agents: Fix custom agent loading when agent files are unparsable
206
207## 1.0.80
208
209- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
210
211## 1.0.77
212
213- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
214- SDK: Add session support and permission denial tracking
215- Fix token limit errors in conversation summarization
216- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
217
218## 1.0.73
219
220- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
221- MCP: Press Esc to cancel OAuth authentication flows
222- Bash: Improved command validation and reduced false security warnings
223- UI: Enhanced spinner animations and status line visual hierarchy
224- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
225
226## 1.0.72
227
228- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
229
230## 1.0.71
231
232- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
233- Customizable status line: add your terminal prompt to Claude Code with /statusline
234
235## 1.0.70
236
237- Performance: Optimized message rendering for better performance with large contexts
238- Windows: Fixed native file search, ripgrep, and subagent functionality
239- Added support for @-mentions in slash command arguments
240
241## 1.0.69
242
243- Upgraded Opus to version 4.1
244
245## 1.0.68
246
247- Fix incorrect model names being used for certain commands like `/pr-comments`
248- 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.
249- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
250- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
251- SDK: Added canUseTool callback support for tool confirmation
252- Added `disableAllHooks` setting
253- Improved file suggestions performance in large repos
254
255## 1.0.65
256
257- IDE: Fixed connection stability issues and error handling for diagnostics
258- Windows: Fixed shell environment setup for users without .bashrc files
259
260## 1.0.64
261
262- Agents: Added model customization support - you can now specify which model an agent should use
263- Agents: Fixed unintended access to the recursive agent tool
264- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
265- SDK: Fixed user input tracking across multi-turn conversations
266- Added hidden files to file search and @-mention suggestions
267
268## 1.0.63
269
270- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
271
272## 1.0.62
273
274- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
275- Hooks: Added SessionStart hook for new session initialization
276- /add-dir command now supports typeahead for directory paths
277- Improved network connectivity check reliability
278
279## 1.0.61
280
281- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
282- Settings: Added `--settings` flag to load settings from a JSON file
283- Settings: Fixed resolution of settings files paths that are symlinks
284- OTEL: Fixed reporting of wrong organization after authentication changes
285- Slash commands: Fixed permissions checking for allowed-tools with Bash
286- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
287- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
288- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
289
290## 1.0.60
291
292- You can now create custom subagents for specialized tasks! Run /agents to get started
293
294## 1.0.59
295
296- SDK: Added tool confirmation support with canUseTool callback
297- SDK: Allow specifying env for spawned process
298- Hooks: Exposed PermissionDecision to hooks (including "ask")
299- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
300- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
301
302## 1.0.58
303
304- Added support for reading PDFs
305- MCP: Improved server health status display in 'claude mcp list'
306- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
307
308## 1.0.57
309
310- Added support for specifying a model in slash commands
311- Improved permission messages to help Claude understand allowed tools
312- Fix: Remove trailing newlines from bash output in terminal wrapping
313
314## 1.0.56
315
316- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
317- Fixes for WSL IDE detection
318- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
319
320## 1.0.55
321
322- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
323- Windows: fixed Ctrl+Z crash
324- SDK: Added ability to capture error logging
325- Add --system-prompt-file option to override system prompt in print mode
326
327## 1.0.54
328
329- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
330- Custom slash commands: Added argument-hint to frontmatter
331- Windows: OAuth uses port 45454 and properly constructs browser URL
332- Windows: mode switching now uses alt + m, and plan mode renders properly
333- Shell: Switch to in-memory shell snapshot to fix file-related errors
334
335## 1.0.53
336
337- Updated @-mention file truncation from 100 lines to 2000 lines
338- 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).
339
340## 1.0.52
341
342- Added support for MCP server instructions
343
344## 1.0.51
345
346- Added support for native Windows (requires Git for Windows)
347- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
348- Settings: /doctor can now help you identify and fix invalid setting files
349- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
350- Increased auto-compact warning threshold from 60% to 80%
351- Fixed an issue with handling user directories with spaces for shell snapshots
352- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
353- Custom slash commands: Fixed user-level commands in subdirectories
354- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
355
356## 1.0.48
357
358- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
359- Added progress messages to Bash tool based on the last 5 lines of command output
360- Added expanding variables support for MCP server configuration
361- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
362- Improved IDE extension path handling when Claude Code runs in WSL
363- Hooks: Added a PreCompact hook
364- Vim mode: Added c, f/F, t/T
365
366## 1.0.45
367
368- Redesigned Search (Grep) tool with new tool input parameters and features
369- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
370- Fixed config file corruption issue by enforcing atomic writes
371- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
372- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
373- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
374
375## 1.0.44
376
377- New /export command lets you quickly export a conversation for sharing
378- MCP: resource_link tool results are now supported
379- MCP: tool annotations and tool titles now display in /mcp view
380- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
381
382## 1.0.43
383
384- Fixed a bug where the theme selector was saving excessively
385- Hooks: Added EPIPE system error handling
386
387## 1.0.42
388
389- Added tilde (`~`) expansion support to `/add-dir` command
390
391## 1.0.41
392
393- Hooks: Split Stop hook triggering into Stop and SubagentStop
394- Hooks: Enabled optional timeout configuration for each command
395- Hooks: Added "hook_event_name" to hook input
396- Fixed a bug where MCP tools would display twice in tool list
397- New tool parameters JSON for Bash tool in `tool_decision` event
398
399## 1.0.40
400
401- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
402
403## 1.0.39
404
405- New Active Time metric in OpenTelemetry logging
406
407## 1.0.38
408
409- 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
410
411## 1.0.37
412
413- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
414
415## 1.0.36
416
417- Web search now takes today's date into context
418- Fixed a bug where stdio MCP servers were not terminating properly on exit
419
420## 1.0.35
421
422- Added support for MCP OAuth Authorization Server discovery
423
424## 1.0.34
425
426- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
427
428## 1.0.33
429
430- Improved logging functionality with session ID support
431- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
432- Improvements to plan mode
433
434## 1.0.32
435
436- Updated loopback config for litellm
437- Added forceLoginMethod setting to bypass login selection screen
438
439## 1.0.31
440
441- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
442
443## 1.0.30
444
445- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
446- Improved file path autocomplete with filename matching
447- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
448- Enhanced jq regex support for complex filters with pipes and select
449
450## 1.0.29
451
452- Improved CJK character support in cursor navigation and rendering
453
454## 1.0.28
455
456- Slash commands: Fix selector display during history navigation
457- Resizes images before upload to prevent API size limit errors
458- Added XDG_CONFIG_HOME support to configuration directory
459- Performance optimizations for memory usage
460- New attributes (terminal.type, language) in OpenTelemetry logging
461
462## 1.0.27
463
464- Streamable HTTP MCP servers are now supported
465- Remote MCP servers (SSE and HTTP) now support OAuth
466- MCP resources can now be @-mentioned
467- /resume slash command to switch conversations within Claude Code
468
469## 1.0.25
470
471- Slash commands: moved "project" and "user" prefixes to descriptions
472- Slash commands: improved reliability for command discovery
473- Improved support for Ghostty
474- Improved web search reliability
475
476## 1.0.24
477
478- Improved /mcp output
479- Fixed a bug where settings arrays got overwritten instead of merged
480
481## 1.0.23
482
483- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
484- Released Python SDK: pip install claude-code-sdk to get started
485
486## 1.0.22
487
488- SDK: Renamed `total_cost` to `total_cost_usd`
489
490## 1.0.21
491
492- Improved editing of files with tab-based indentation
493- Fix for tool_use without matching tool_result errors
494- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
495
496## 1.0.18
497
498- Added --add-dir CLI argument for specifying additional working directories
499- Added streaming input support without require -p flag
500- Improved startup performance and session storage performance
501- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
502- Added detailed MCP server tools display (/mcp)
503- MCP authentication and permission improvements
504- Added auto-reconnection for MCP SSE connections on disconnect
505- Fixed issue where pasted content was lost when dialogs appeared
506
507## 1.0.17
508
509- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
510- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
511- MCP server list UI improvements
512- Update Claude Code process title to display "claude" instead of "node"
513
514## 1.0.11
515
516- Claude Code can now also be used with a Claude Pro subscription
517- Added /upgrade for smoother switching to Claude Max plans
518- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
519- Improved shell configuration error handling
520- Improved todo list handling during compaction
521
522## 1.0.10
523
524- Added markdown table support
525- Improved streaming performance
526
527## 1.0.8
528
529- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
530- Increased default otel interval from 1s -> 5s
531- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
532- Fixed a regression where search tools unnecessarily asked for permissions
533- Added support for triggering thinking non-English languages
534- Improved compacting UI
535
536## 1.0.7
537
538- Renamed /allowed-tools -> /permissions
539- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
540- Deprecated claude config commands in favor of editing settings.json
541- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
542- Improved error handling for /install-github-app
543- Bugfixes, UI polish, and tool reliability improvements
544
545## 1.0.6
546
547- Improved edit reliability for tab-indented files
548- Respect CLAUDE_CONFIG_DIR everywhere
549- Reduced unnecessary tool permission prompts
550- Added support for symlinks in @file typeahead
551- Bugfixes, UI polish, and tool reliability improvements
552
553## 1.0.4
554
555- Fixed a bug where MCP tool errors weren't being parsed correctly
556
557## 1.0.1
558
559- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
560- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
561- Updated documentation links and OAuth process descriptions
562
563## 1.0.0
564
565- Claude Code is now generally available
566- Introducing Sonnet 4 and Opus 4 models
567
568## 0.2.125
569
570- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
571- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
572
573## 0.2.117
574
575- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
576- Introduced settings.cleanupPeriodDays
577- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
578- Introduced --debug mode
579
580## 0.2.108
581
582- You can now send messages to Claude while it works to steer Claude in real-time
583- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
584- Fixed a bug where thinking was not working in -p mode
585- Fixed a regression in /cost reporting
586- Deprecated MCP wizard interface in favor of other MCP commands
587- Lots of other bugfixes and improvements
588
589## 0.2.107
590
591- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
592
593## 0.2.106
594
595- MCP SSE server configs can now specify custom headers
596- Fixed a bug where MCP permission prompt didn't always show correctly
597
598## 0.2.105
599
600- Claude can now search the web
601- Moved system & account status to /status
602- Added word movement keybindings for Vim
603- Improved latency for startup, todo tool, and file edits
604
605## 0.2.102
606
607- Improved thinking triggering reliability
608- Improved @mention reliability for images and folders
609- You can now paste multiple large chunks into one prompt
610
611## 0.2.100
612
613- Fixed a crash caused by a stack overflow error
614- Made db storage optional; missing db support disables --continue and --resume
615
616## 0.2.98
617
618- Fixed an issue where auto-compact was running twice
619
620## 0.2.96
621
622- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
623
624## 0.2.93
625
626- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
627- Claude now has access to a Todo list that helps it stay on track and be more organized
628
629## 0.2.82
630
631- Added support for --disallowedTools
632- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
633
634## 0.2.75
635
636- Hit Enter to queue up additional messages while Claude is working
637- Drag in or copy/paste image files directly into the prompt
638- @-mention files to directly add them to context
639- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
640- Improved performance for filename auto-complete
641
642## 0.2.74
643
644- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
645- Task tool can now perform writes and run bash commands
646
647## 0.2.72
648
649- Updated spinner to indicate tokens loaded and tool usage
650
651## 0.2.70
652
653- Network commands like curl are now available for Claude to use
654- Claude can now run multiple web queries in parallel
655- Pressing ESC once immediately interrupts Claude in Auto-accept mode
656
657## 0.2.69
658
659- Fixed UI glitches with improved Select component behavior
660- Enhanced terminal output display with better text truncation logic
661
662## 0.2.67
663
664- Shared project permission rules can be saved in .claude/settings.json
665
666## 0.2.66
667
668- Print mode (-p) now supports streaming output via --output-format=stream-json
669- Fixed issue where pasting could trigger memory or bash mode unexpectedly
670
671## 0.2.63
672
673- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
674
675## 0.2.61
676
677- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
678- Enhanced image detection for more reliable clipboard paste functionality
679- Fixed an issue where ESC key could crash the conversation history selector
680
681## 0.2.59
682
683- Copy+paste images directly into your prompt
684- Improved progress indicators for bash and fetch tools
685- Bugfixes for non-interactive mode (-p)
686
687## 0.2.54
688
689- Quickly add to Memory by starting your message with '#'
690- Press ctrl+r to see full output for long tool results
691- Added support for MCP SSE transport
692
693## 0.2.53
694
695- New web fetch tool lets Claude view URLs that you paste in
696- Fixed a bug with JPEG detection
697
698## 0.2.50
699
700- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
701
702## 0.2.49
703
704- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
705
706## 0.2.47
707
708- Press Tab to auto-complete file and folder names
709- Press Shift + Tab to toggle auto-accept for file edits
710- Automatic conversation compaction for infinite conversation length (toggle with /config)
711
712## 0.2.44
713
714- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
715
716## 0.2.41
717
718- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
719- MCP server startup no longer blocks the app from starting up
720
721## 0.2.37
722
723- New /release-notes command lets you view release notes at any time
724- `claude config add/remove` commands now accept multiple values separated by commas or spaces
725
726## 0.2.36
727
728- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
729- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
730
731## 0.2.34
732
733- Vim bindings for text input - enable with /vim or /config
734
735## 0.2.32
736
737- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
738- Fix for some PersistentShell issues
739
740## 0.2.31
741
742- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
743- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
744
745## 0.2.30
746
747- Added ANSI color theme for better terminal compatibility
748- Fixed issue where slash command arguments weren't being sent properly
749- (Mac-only) API keys are now stored in macOS Keychain
750
751## 0.2.26
752
753- New /approved-tools command for managing tool permissions
754- Word-level diff display for improved code readability
755- Fuzzy matching for slash commands
756
757## 0.2.21
758
759- Fuzzy matching for /commands