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 af073ad · on Oct 9, 2025 History
773 lines · 27.8 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.12
4
5- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
6- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
7- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
8- `/plugin validate` command for validating plugin structure and configuration
9- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
10- Plugin documentation available at https://docs.claude.com/en/docs/claude-code/plugins
11- Comprehensive error messages and diagnostics via `/doctor` command
12- Avoid flickering in `/model` selector
13- Improvements to `/help`
14- Avoid mentioning hooks in `/resume` summaries
15- Changes to the "verbose" setting in `/config` now persist across sessions
16
17## 2.0.11
18
19- Reduced system prompt size by 1.4k tokens
20- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
21- Fixed Opus fallback rate limit errors appearing incorrectly
22- Fixed /add-dir command selecting wrong default tab
23
24## 2.0.10
25
26- Rewrote terminal renderer for buttery smooth UI
27- Enable/disable MCP servers by @mentioning, or in /mcp
28- Added tab completion for shell commands in bash mode
29- PreToolUse hooks can now modify tool inputs
30- Press Ctrl-G to edit your prompt in your system's configured text editor
31- Fixes for bash permission checks with environment variables in the command
32
33## 2.0.9
34
35- Fix regression where bash backgrounding stopped working
36
37## 2.0.8
38
39- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
40- IDE: Add drag-and-drop support for files and folders in chat
41- /context: Fix counting for thinking blocks
42- Improve message rendering for users with light themes on dark terminals
43- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
44
45## 2.0.5
46
47- IDE: Fix IME unintended message submission with Enter and Tab
48- IDE: Add "Open in Terminal" link in login screen
49- Fix unhandled OAuth expiration 401 API errors
50- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
51
52## 2.0.1
53
54- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
55- Various bug fixes and presentation improvements
56
57## 2.0.0
58
59- New native VS Code extension
60- Fresh coat of paint throughout the whole app
61- /rewind a conversation to undo code changes
62- /usage command to see plan limits
63- Tab to toggle thinking (sticky across sessions)
64- Ctrl-R to search history
65- Unshipped claude config command
66- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
67- SDK: The Claude Code SDK is now the Claude Agent SDK
68- Add subagents dynamically with `--agents` flag
69
70## 1.0.126
71
72- Enable /context command for Bedrock and Vertex
73- Add mTLS support for HTTP-based OpenTelemetry exporters
74
75## 1.0.124
76
77- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
78- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
79- No longer inform Claude of the list of allowed tools when permission is denied
80- Fixed security vulnerability in Bash tool permission checks
81- Improved VSCode extension performance for large files
82
83## 1.0.123
84
85- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
86- Fixed thinking mode triggering on negation phrases like "don't think"
87- Fixed rendering performance degradation during token streaming
88- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
89- Enhanced BashTool environment snapshot logging
90- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
91- Migrated --debug logging to a file, to enable easy tailing & filtering
92
93## 1.0.120
94
95- Fix input lag during typing, especially noticeable with large prompts
96- Improved VSCode extension command registry and sessions dialog user experience
97- Enhanced sessions dialog responsiveness and visual feedback
98- Fixed IDE compatibility issue by removing worktree support check
99- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
100
101## 1.0.119
102
103- Fix Windows issue where process visually freezes on entering interactive mode
104- Support dynamic headers for MCP servers via headersHelper configuration
105- Fix thinking mode not working in headless sessions
106- Fix slash commands now properly update allowed tools instead of replacing them
107
108## 1.0.117
109
110- Add Ctrl-R history search to recall previous commands like bash/zsh
111- Fix input lag while typing, especially on Windows
112- Add sed command to auto-allowed commands in acceptEdits mode
113- Fix Windows PATH comparison to be case-insensitive for drive letters
114- Add permissions management hint to /add-dir output
115
116## 1.0.115
117
118- Improve thinking mode display with enhanced visual effects
119- Type /t to temporarily disable thinking mode in your prompt
120- Improve path validation for glob and grep tools
121- Show condensed output for post-tool hooks to reduce visual clutter
122- Fix visual feedback when loading state completes
123- Improve UI consistency for permission request dialogs
124
125## 1.0.113
126
127- Deprecated piped input in interactive mode
128- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
129
130## 1.0.112
131
132- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
133- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
134- Hooks: Added systemMessage support for SessionEnd hooks
135- Added `spinnerTipsEnabled` setting to disable spinner tips
136- IDE: Various improvements and bug fixes
137
138## 1.0.111
139
140- /model now validates provided model names
141- Fixed Bash tool crashes caused by malformed shell syntax parsing
142
143## 1.0.110
144
145- /terminal-setup command now supports WezTerm
146- MCP: OAuth tokens now proactively refresh before expiration
147- Fixed reliability issues with background Bash processes
148
149## 1.0.109
150
151- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
152
153## 1.0.106
154
155- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
156
157## 1.0.97
158
159- Settings: /doctor now validates permission rule syntax and suggests corrections
160
161## 1.0.94
162
163- Vertex: add support for global endpoints for supported models
164- /memory command now allows direct editing of all imported memory files
165- SDK: Add custom tools as callbacks
166- Added /todos command to list current todo items
167
168## 1.0.93
169
170- Windows: Add alt + v shortcut for pasting images from clipboard
171- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
172
173## 1.0.90
174
175- Settings file changes take effect immediately - no restart required
176
177## 1.0.88
178
179- Fixed issue causing "OAuth authentication is currently not supported"
180- Status line input now includes `exceeds_200k_tokens`
181- Fixed incorrect usage tracking in /cost.
182- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
183- Bedrock: Updated default Sonnet model to Sonnet 4
184
185## 1.0.86
186
187- Added /context to help users self-serve debug context issues
188- SDK: Added UUID support for all SDK messages
189- SDK: Added `--replay-user-messages` to replay user messages back to stdout
190
191## 1.0.85
192
193- Status line input now includes session cost info
194- Hooks: Introduced SessionEnd hook
195
196## 1.0.84
197
198- Fix tool_use/tool_result id mismatch error when network is unstable
199- Fix Claude sometimes ignoring real-time steering when wrapping up a task
200- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
201- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
202
203## 1.0.83
204
205- @-mention: Support files with spaces in path
206- New shimmering spinner
207
208## 1.0.82
209
210- SDK: Add request cancellation support
211- SDK: New additionalDirectories option to search custom paths, improved slash command processing
212- Settings: Validation prevents invalid fields in .claude/settings.json files
213- MCP: Improve tool name consistency
214- Bash: Fix crash when Claude tries to automatically read large files
215
216## 1.0.81
217
218- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
219- Agents: Fix custom agent loading when agent files are unparsable
220
221## 1.0.80
222
223- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
224
225## 1.0.77
226
227- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
228- SDK: Add session support and permission denial tracking
229- Fix token limit errors in conversation summarization
230- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
231
232## 1.0.73
233
234- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
235- MCP: Press Esc to cancel OAuth authentication flows
236- Bash: Improved command validation and reduced false security warnings
237- UI: Enhanced spinner animations and status line visual hierarchy
238- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
239
240## 1.0.72
241
242- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
243
244## 1.0.71
245
246- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
247- Customizable status line: add your terminal prompt to Claude Code with /statusline
248
249## 1.0.70
250
251- Performance: Optimized message rendering for better performance with large contexts
252- Windows: Fixed native file search, ripgrep, and subagent functionality
253- Added support for @-mentions in slash command arguments
254
255## 1.0.69
256
257- Upgraded Opus to version 4.1
258
259## 1.0.68
260
261- Fix incorrect model names being used for certain commands like `/pr-comments`
262- 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.
263- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
264- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
265- SDK: Added canUseTool callback support for tool confirmation
266- Added `disableAllHooks` setting
267- Improved file suggestions performance in large repos
268
269## 1.0.65
270
271- IDE: Fixed connection stability issues and error handling for diagnostics
272- Windows: Fixed shell environment setup for users without .bashrc files
273
274## 1.0.64
275
276- Agents: Added model customization support - you can now specify which model an agent should use
277- Agents: Fixed unintended access to the recursive agent tool
278- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
279- SDK: Fixed user input tracking across multi-turn conversations
280- Added hidden files to file search and @-mention suggestions
281
282## 1.0.63
283
284- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
285
286## 1.0.62
287
288- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
289- Hooks: Added SessionStart hook for new session initialization
290- /add-dir command now supports typeahead for directory paths
291- Improved network connectivity check reliability
292
293## 1.0.61
294
295- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
296- Settings: Added `--settings` flag to load settings from a JSON file
297- Settings: Fixed resolution of settings files paths that are symlinks
298- OTEL: Fixed reporting of wrong organization after authentication changes
299- Slash commands: Fixed permissions checking for allowed-tools with Bash
300- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
301- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
302- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
303
304## 1.0.60
305
306- You can now create custom subagents for specialized tasks! Run /agents to get started
307
308## 1.0.59
309
310- SDK: Added tool confirmation support with canUseTool callback
311- SDK: Allow specifying env for spawned process
312- Hooks: Exposed PermissionDecision to hooks (including "ask")
313- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
314- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
315
316## 1.0.58
317
318- Added support for reading PDFs
319- MCP: Improved server health status display in 'claude mcp list'
320- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
321
322## 1.0.57
323
324- Added support for specifying a model in slash commands
325- Improved permission messages to help Claude understand allowed tools
326- Fix: Remove trailing newlines from bash output in terminal wrapping
327
328## 1.0.56
329
330- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
331- Fixes for WSL IDE detection
332- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
333
334## 1.0.55
335
336- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
337- Windows: fixed Ctrl+Z crash
338- SDK: Added ability to capture error logging
339- Add --system-prompt-file option to override system prompt in print mode
340
341## 1.0.54
342
343- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
344- Custom slash commands: Added argument-hint to frontmatter
345- Windows: OAuth uses port 45454 and properly constructs browser URL
346- Windows: mode switching now uses alt + m, and plan mode renders properly
347- Shell: Switch to in-memory shell snapshot to fix file-related errors
348
349## 1.0.53
350
351- Updated @-mention file truncation from 100 lines to 2000 lines
352- 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).
353
354## 1.0.52
355
356- Added support for MCP server instructions
357
358## 1.0.51
359
360- Added support for native Windows (requires Git for Windows)
361- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
362- Settings: /doctor can now help you identify and fix invalid setting files
363- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
364- Increased auto-compact warning threshold from 60% to 80%
365- Fixed an issue with handling user directories with spaces for shell snapshots
366- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
367- Custom slash commands: Fixed user-level commands in subdirectories
368- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
369
370## 1.0.48
371
372- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
373- Added progress messages to Bash tool based on the last 5 lines of command output
374- Added expanding variables support for MCP server configuration
375- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
376- Improved IDE extension path handling when Claude Code runs in WSL
377- Hooks: Added a PreCompact hook
378- Vim mode: Added c, f/F, t/T
379
380## 1.0.45
381
382- Redesigned Search (Grep) tool with new tool input parameters and features
383- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
384- Fixed config file corruption issue by enforcing atomic writes
385- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
386- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
387- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
388
389## 1.0.44
390
391- New /export command lets you quickly export a conversation for sharing
392- MCP: resource_link tool results are now supported
393- MCP: tool annotations and tool titles now display in /mcp view
394- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
395
396## 1.0.43
397
398- Fixed a bug where the theme selector was saving excessively
399- Hooks: Added EPIPE system error handling
400
401## 1.0.42
402
403- Added tilde (`~`) expansion support to `/add-dir` command
404
405## 1.0.41
406
407- Hooks: Split Stop hook triggering into Stop and SubagentStop
408- Hooks: Enabled optional timeout configuration for each command
409- Hooks: Added "hook_event_name" to hook input
410- Fixed a bug where MCP tools would display twice in tool list
411- New tool parameters JSON for Bash tool in `tool_decision` event
412
413## 1.0.40
414
415- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
416
417## 1.0.39
418
419- New Active Time metric in OpenTelemetry logging
420
421## 1.0.38
422
423- 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
424
425## 1.0.37
426
427- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
428
429## 1.0.36
430
431- Web search now takes today's date into context
432- Fixed a bug where stdio MCP servers were not terminating properly on exit
433
434## 1.0.35
435
436- Added support for MCP OAuth Authorization Server discovery
437
438## 1.0.34
439
440- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
441
442## 1.0.33
443
444- Improved logging functionality with session ID support
445- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
446- Improvements to plan mode
447
448## 1.0.32
449
450- Updated loopback config for litellm
451- Added forceLoginMethod setting to bypass login selection screen
452
453## 1.0.31
454
455- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
456
457## 1.0.30
458
459- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
460- Improved file path autocomplete with filename matching
461- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
462- Enhanced jq regex support for complex filters with pipes and select
463
464## 1.0.29
465
466- Improved CJK character support in cursor navigation and rendering
467
468## 1.0.28
469
470- Slash commands: Fix selector display during history navigation
471- Resizes images before upload to prevent API size limit errors
472- Added XDG_CONFIG_HOME support to configuration directory
473- Performance optimizations for memory usage
474- New attributes (terminal.type, language) in OpenTelemetry logging
475
476## 1.0.27
477
478- Streamable HTTP MCP servers are now supported
479- Remote MCP servers (SSE and HTTP) now support OAuth
480- MCP resources can now be @-mentioned
481- /resume slash command to switch conversations within Claude Code
482
483## 1.0.25
484
485- Slash commands: moved "project" and "user" prefixes to descriptions
486- Slash commands: improved reliability for command discovery
487- Improved support for Ghostty
488- Improved web search reliability
489
490## 1.0.24
491
492- Improved /mcp output
493- Fixed a bug where settings arrays got overwritten instead of merged
494
495## 1.0.23
496
497- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
498- Released Python SDK: pip install claude-code-sdk to get started
499
500## 1.0.22
501
502- SDK: Renamed `total_cost` to `total_cost_usd`
503
504## 1.0.21
505
506- Improved editing of files with tab-based indentation
507- Fix for tool_use without matching tool_result errors
508- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
509
510## 1.0.18
511
512- Added --add-dir CLI argument for specifying additional working directories
513- Added streaming input support without require -p flag
514- Improved startup performance and session storage performance
515- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
516- Added detailed MCP server tools display (/mcp)
517- MCP authentication and permission improvements
518- Added auto-reconnection for MCP SSE connections on disconnect
519- Fixed issue where pasted content was lost when dialogs appeared
520
521## 1.0.17
522
523- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
524- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
525- MCP server list UI improvements
526- Update Claude Code process title to display "claude" instead of "node"
527
528## 1.0.11
529
530- Claude Code can now also be used with a Claude Pro subscription
531- Added /upgrade for smoother switching to Claude Max plans
532- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
533- Improved shell configuration error handling
534- Improved todo list handling during compaction
535
536## 1.0.10
537
538- Added markdown table support
539- Improved streaming performance
540
541## 1.0.8
542
543- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
544- Increased default otel interval from 1s -> 5s
545- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
546- Fixed a regression where search tools unnecessarily asked for permissions
547- Added support for triggering thinking non-English languages
548- Improved compacting UI
549
550## 1.0.7
551
552- Renamed /allowed-tools -> /permissions
553- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
554- Deprecated claude config commands in favor of editing settings.json
555- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
556- Improved error handling for /install-github-app
557- Bugfixes, UI polish, and tool reliability improvements
558
559## 1.0.6
560
561- Improved edit reliability for tab-indented files
562- Respect CLAUDE_CONFIG_DIR everywhere
563- Reduced unnecessary tool permission prompts
564- Added support for symlinks in @file typeahead
565- Bugfixes, UI polish, and tool reliability improvements
566
567## 1.0.4
568
569- Fixed a bug where MCP tool errors weren't being parsed correctly
570
571## 1.0.1
572
573- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
574- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
575- Updated documentation links and OAuth process descriptions
576
577## 1.0.0
578
579- Claude Code is now generally available
580- Introducing Sonnet 4 and Opus 4 models
581
582## 0.2.125
583
584- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
585- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
586
587## 0.2.117
588
589- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
590- Introduced settings.cleanupPeriodDays
591- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
592- Introduced --debug mode
593
594## 0.2.108
595
596- You can now send messages to Claude while it works to steer Claude in real-time
597- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
598- Fixed a bug where thinking was not working in -p mode
599- Fixed a regression in /cost reporting
600- Deprecated MCP wizard interface in favor of other MCP commands
601- Lots of other bugfixes and improvements
602
603## 0.2.107
604
605- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
606
607## 0.2.106
608
609- MCP SSE server configs can now specify custom headers
610- Fixed a bug where MCP permission prompt didn't always show correctly
611
612## 0.2.105
613
614- Claude can now search the web
615- Moved system & account status to /status
616- Added word movement keybindings for Vim
617- Improved latency for startup, todo tool, and file edits
618
619## 0.2.102
620
621- Improved thinking triggering reliability
622- Improved @mention reliability for images and folders
623- You can now paste multiple large chunks into one prompt
624
625## 0.2.100
626
627- Fixed a crash caused by a stack overflow error
628- Made db storage optional; missing db support disables --continue and --resume
629
630## 0.2.98
631
632- Fixed an issue where auto-compact was running twice
633
634## 0.2.96
635
636- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
637
638## 0.2.93
639
640- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
641- Claude now has access to a Todo list that helps it stay on track and be more organized
642
643## 0.2.82
644
645- Added support for --disallowedTools
646- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
647
648## 0.2.75
649
650- Hit Enter to queue up additional messages while Claude is working
651- Drag in or copy/paste image files directly into the prompt
652- @-mention files to directly add them to context
653- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
654- Improved performance for filename auto-complete
655
656## 0.2.74
657
658- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
659- Task tool can now perform writes and run bash commands
660
661## 0.2.72
662
663- Updated spinner to indicate tokens loaded and tool usage
664
665## 0.2.70
666
667- Network commands like curl are now available for Claude to use
668- Claude can now run multiple web queries in parallel
669- Pressing ESC once immediately interrupts Claude in Auto-accept mode
670
671## 0.2.69
672
673- Fixed UI glitches with improved Select component behavior
674- Enhanced terminal output display with better text truncation logic
675
676## 0.2.67
677
678- Shared project permission rules can be saved in .claude/settings.json
679
680## 0.2.66
681
682- Print mode (-p) now supports streaming output via --output-format=stream-json
683- Fixed issue where pasting could trigger memory or bash mode unexpectedly
684
685## 0.2.63
686
687- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
688
689## 0.2.61
690
691- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
692- Enhanced image detection for more reliable clipboard paste functionality
693- Fixed an issue where ESC key could crash the conversation history selector
694
695## 0.2.59
696
697- Copy+paste images directly into your prompt
698- Improved progress indicators for bash and fetch tools
699- Bugfixes for non-interactive mode (-p)
700
701## 0.2.54
702
703- Quickly add to Memory by starting your message with '#'
704- Press ctrl+r to see full output for long tool results
705- Added support for MCP SSE transport
706
707## 0.2.53
708
709- New web fetch tool lets Claude view URLs that you paste in
710- Fixed a bug with JPEG detection
711
712## 0.2.50
713
714- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
715
716## 0.2.49
717
718- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
719
720## 0.2.47
721
722- Press Tab to auto-complete file and folder names
723- Press Shift + Tab to toggle auto-accept for file edits
724- Automatic conversation compaction for infinite conversation length (toggle with /config)
725
726## 0.2.44
727
728- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
729
730## 0.2.41
731
732- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
733- MCP server startup no longer blocks the app from starting up
734
735## 0.2.37
736
737- New /release-notes command lets you view release notes at any time
738- `claude config add/remove` commands now accept multiple values separated by commas or spaces
739
740## 0.2.36
741
742- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
743- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
744
745## 0.2.34
746
747- Vim bindings for text input - enable with /vim or /config
748
749## 0.2.32
750
751- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
752- Fix for some PersistentShell issues
753
754## 0.2.31
755
756- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
757- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
758
759## 0.2.30
760
761- Added ANSI color theme for better terminal compatibility
762- Fixed issue where slash command arguments weren't being sent properly
763- (Mac-only) API keys are now stored in macOS Keychain
764
765## 0.2.26
766
767- New /approved-tools command for managing tool permissions
768- Word-level diff display for improved code readability
769- Fuzzy matching for slash commands
770
771## 0.2.21
772
773- Fuzzy matching for /commands