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 6880bcb · on Dec 17, 2025 History
1,151 lines · 47.4 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.72
4
5- Added Claude in Chrome (Beta) feature that works with the Chrome extension (https://claude.ai/chrome) to let you control your browser directly from Claude Code
6- Reduced terminal flickering
7- Added scannable QR code to mobile app tip for quick app downloads
8- Added loading indicator when resuming conversations for better feedback
9- Fixed `/context` command not respecting custom system prompts in non-interactive mode
10- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
11- Improved @ mention file suggestion speed (~3x faster in git repositories)
12- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
13- Improved settings validation errors to be more prominent
14- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
15
16## 2.0.71
17
18- Added /config toggle to enable/disable prompt suggestions
19- Added `/settings` as an alias for the `/config` command
20- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
21- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
22- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
23- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
24- New syntax highlighting engine for native build
25
26## 2.0.70
27
28- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
29- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
30- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
31- Added `plan_mode_required` spawn parameter for teammates to require plan approval before implementing changes
32- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
33- Fixed input being cleared when processing queued commands while the user was typing
34- Fixed prompt suggestions replacing typed input when pressing Tab
35- Fixed diff view not updating when terminal is resized
36- Improved memory usage by 3x for large conversations
37- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
38- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
39- Fix thinking mode toggle in /config not persisting correctly
40- Improve UI for file creation permission dialog
41
42## 2.0.69
43
44- Minor bugfixes
45
46## 2.0.68
47
48- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
49- Fixed a bug where disallowed MCP tools were visible to the model
50- Fixed an issue where steering messages could be lost while a subagent is working
51- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
52- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
53- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
54
55## 2.0.67
56
57- Thinking mode is now enabled by default for Opus 4.5
58- Thinking mode configuration has moved to /config
59- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
60- Show reason why autoupdater is disabled in `/doctor`
61- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
62- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
63- Fixed scroll position resetting after deleting a permission rule in `/permissions`
64- Fixed word deletion (opt+delete) and word navigation (opt+arrow) not working correctly with non-Latin text such as Cyrillic, Greek, Arabic, Hebrew, Thai, and Chinese
65- Fixed `claude install --force` not bypassing stale lock files
66- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
67- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
68
69## 2.0.65
70
71- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
72- Added context window information to status line input
73- Added `fileSuggestion` setting for custom `@` file search commands
74- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
75- Fixed prompt not being saved to history when aborting a query with Escape
76- Fixed Read tool image handling to identify format from bytes instead of file extension
77
78## 2.0.64
79
80- Made auto-compacting instant
81- Agents and bash commands can run asynchronously and send messages to wake up the main agent
82- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
83- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
84- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
85- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
86- Fixed auto-loading .env when using native installer
87- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
88- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
89- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
90- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
91- Bedrock: Improve efficiency of token counting
92- Bedrock: Add support for `aws login` AWS Management Console credentials
93- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
94
95## 2.0.62
96
97- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
98- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
99- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
100- Fixed slash command selection not working when multiple commands share the same name
101- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
102- Fixed running versions getting removed because lock file incorrectly going stale
103- Fixed IDE diff tab not closing when rejecting file changes
104
105## 2.0.61
106
107- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
108
109## 2.0.60
110
111- Added background agent support. Agents run in the background while you work
112- Added --disable-slash-commands CLI flag to disable all slash commands
113- Added model name to "Co-Authored-By" commit messages
114- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
115- Updated Fetch to skip summarization for pre-approved websites
116- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
117
118## 2.0.59
119
120- Added --agent CLI flag to override the agent setting for the current session
121- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
122- VS Code: Fixed .claude.json config file being read from incorrect location
123
124## 2.0.58
125
126- Pro users now have access to Opus 4.5 as part of their subscription!
127- Fixed timer duration showing "11m 60s" instead of "12m 0s"
128- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
129
130## 2.0.57
131
132- Added feedback input when rejecting plans, allowing users to tell Claude what to change
133- VSCode: Added streaming message support for real-time response display
134
135## 2.0.56
136
137- Added setting to enable/disable terminal progress bar (OSC 9;4)
138- VSCode Extension: Added support for VS Code's secondary sidebar (VS Code 1.97+), allowing Claude Code to be displayed in the right sidebar while keeping the file explorer on the left. Requires setting sidebar as Preferred Location in the config.
139
140## 2.0.55
141
142- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
143- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
144- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
145- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
146
147## 2.0.54
148
149- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
150- Fix issue with excessive iTerm notifications
151
152## 2.0.52
153
154- Fixed duplicate message display when starting Claude with a command line argument
155- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
156- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
157- Permit some uses of `$!` in bash commands
158
159## 2.0.51
160
161- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
162- Introducing Claude Code for Desktop: https://claude.com/download
163- To give you room to try out our new model, we've updated usage limits for Claude Code users. See the Claude Opus 4.5 blog for full details
164- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
165- Plan Mode now builds more precise plans and executes more thoroughly
166- Usage limit notifications now easier to understand
167- Switched `/usage` back to "% used"
168- Fixed handling of thinking errors
169- Fixed performance regression
170
171## 2.0.50
172
173- Fixed bug preventing calling MCP tools that have nested references in their input schemas
174- Silenced a noisy but harmless error during upgrades
175- Improved ultrathink text display
176- Improved clarity of 5-hour session limit warning message
177
178## 2.0.49
179
180- Added readline-style ctrl-y for pasting deleted text
181- Improved clarity of usage limit warning message
182- Fixed handling of subagent permissions
183
184## 2.0.47
185
186- Improved error messages and validation for `claude --teleport`
187- Improved error handling in `/usage`
188- Fixed race condition with history entry not getting logged at exit
189- Fixed Vertex AI configuration not being applied from `settings.json`
190
191## 2.0.46
192
193- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
194
195## 2.0.45
196
197- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
198- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
199- Send background tasks to Claude Code on the web by starting a message with `&`
200
201## 2.0.43
202
203- Added `permissionMode` field for custom agents
204- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
205- Added skills frontmatter field to declare skills to auto-load for subagents
206- Added the `SubagentStart` hook event
207- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
208- Fixed duplicate rendering of some messages in the UI
209- Fixed some visual flickers
210- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
211
212## 2.0.42
213
214- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
215
216## 2.0.41
217
218- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
219- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
220- Fixed incorrect labeling of user settings vs project settings in command descriptions
221- Fixed crash when plugin command hooks timeout during execution
222- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
223- Fixed broken security documentation links in trust dialogs and onboarding
224- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
225- ctrl-r history search landing on a slash command no longer cancels the search
226- SDK: Support custom timeouts for hooks
227- Allow more safe git commands to run without approval
228- Plugins: Added support for sharing and installing output styles
229- Teleporting a session from web will automatically set the upstream branch
230
231## 2.0.37
232
233- Fixed how idleness is computed for notifications
234- Hooks: Added matcher values for Notification hook events
235- Output Styles: Added `keep-coding-instructions` option to frontmatter
236
237## 2.0.36
238
239- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
240- Fixed queued messages being incorrectly executed as bash commands
241- Fixed input being lost when typing while a queued message is processed
242
243## 2.0.35
244
245- Improve fuzzy search results when searching commands
246- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
247- Added `CLAUDE_CODE_EXIT_AFTER_STOP_DELAY` environment variable to automatically exit SDK mode after a specified idle duration, useful for automated workflows and scripts
248- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
249- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
250
251## 2.0.34
252
253- VSCode Extension: Added setting to configure the initial permission mode for new conversations
254- Improved file path suggestion performance with native Rust-based fuzzy finder
255- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
256- Fixed memory crash when reading or writing large files (especially base64-encoded images)
257
258## 2.0.33
259
260- Native binary installs now launch quicker.
261- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
262- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
263
264## 2.0.32
265
266- Un-deprecate output styles based on community feedback
267- Added `companyAnnouncements` setting for displaying announcements on startup
268- Fixed hook progress messages not updating correctly during PostToolUse hook execution
269
270## 2.0.31
271
272- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
273- Vertex: add support for Web Search on supported models
274- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
275- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
276- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
277- Fixed plugin uninstall not removing plugins
278
279## 2.0.30
280
281- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
282- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
283- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
284- Added prompt-based stop hooks
285- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
286- Enabled SSE MCP servers on native build
287- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
288- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
289- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
290- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
291- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
292- Fixed bug that saved session permissions to local settings
293- Fixed MCP tools not being available to sub-agents
294- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
295- Fixed delay when navigating through typeahead suggestions with arrow keys
296- VSCode: Restored selection indicator in input footer showing current file or code selection status
297
298## 2.0.28
299
300- Plan mode: introduced new Plan subagent
301- Subagents: claude can now choose to resume subagents
302- Subagents: claude can dynamically choose the model used by its subagents
303- SDK: added --max-budget-usd flag
304- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
305- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
306- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
307- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
308- Various other bug fixes
309
310## 2.0.27
311
312- New UI for permission prompts
313- Added current branch filtering and search to session resume screen for easier navigation
314- Fixed directory @-mention causing "No assistant message found" error
315- VSCode Extension: Add config setting to include .gitignored files in file searches
316- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
317
318## 2.0.25
319
320- Removed legacy SDK entrypoint. Please migrate to @anthropic-ai/claude-agent-sdk for future SDK updates: https://platform.claude.com/docs/en/agent-sdk/migration-guide
321
322## 2.0.24
323
324- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
325- Claude Code Web: Support for Web -> CLI teleport
326- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
327- Bedrock: Display awsAuthRefresh output when auth is required
328
329## 2.0.22
330
331- Fixed content layout shift when scrolling through slash commands
332- IDE: Add toggle to enable/disable thinking.
333- Fix bug causing duplicate permission prompts with parallel tool calls
334- Add support for enterprise managed MCP allowlist and denylist
335
336## 2.0.21
337
338- Support MCP `structuredContent` field in tool responses
339- Added an interactive question tool
340- Claude will now ask you questions more often in plan mode
341- Added Haiku 4.5 as a model option for Pro users
342- Fixed an issue where queued commands don't have access to previous messages' output
343
344## 2.0.20
345
346- Added support for Claude Skills
347
348## 2.0.19
349
350- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
351- Fixed a bug where Haiku was unnecessarily called in print mode
352
353## 2.0.17
354
355- Added Haiku 4.5 to model selector!
356- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
357- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
358- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
359- OTEL: support HTTP_PROXY and HTTPS_PROXY
360- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
361
362## 2.0.15
363
364- Fixed bug with resuming where previously created files needed to be read again before writing
365- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
366
367## 2.0.14
368
369- Fix @-mentioning MCP servers to toggle them on/off
370- Improve permission checks for bash with inline env vars
371- Fix ultrathink + thinking toggle
372- Reduce unnecessary logins
373- Document --system-prompt
374- Several improvements to rendering
375- Plugins UI polish
376
377## 2.0.13
378
379- Fixed `/plugin` not working on native build
380
381## 2.0.12
382
383- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
384- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
385- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
386- `/plugin validate` command for validating plugin structure and configuration
387- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
388- Plugin documentation available at https://code.claude.com/docs/en/plugins
389- Comprehensive error messages and diagnostics via `/doctor` command
390- Avoid flickering in `/model` selector
391- Improvements to `/help`
392- Avoid mentioning hooks in `/resume` summaries
393- Changes to the "verbose" setting in `/config` now persist across sessions
394
395## 2.0.11
396
397- Reduced system prompt size by 1.4k tokens
398- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
399- Fixed Opus fallback rate limit errors appearing incorrectly
400- Fixed /add-dir command selecting wrong default tab
401
402## 2.0.10
403
404- Rewrote terminal renderer for buttery smooth UI
405- Enable/disable MCP servers by @mentioning, or in /mcp
406- Added tab completion for shell commands in bash mode
407- PreToolUse hooks can now modify tool inputs
408- Press Ctrl-G to edit your prompt in your system's configured text editor
409- Fixes for bash permission checks with environment variables in the command
410
411## 2.0.9
412
413- Fix regression where bash backgrounding stopped working
414
415## 2.0.8
416
417- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
418- IDE: Add drag-and-drop support for files and folders in chat
419- /context: Fix counting for thinking blocks
420- Improve message rendering for users with light themes on dark terminals
421- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
422
423## 2.0.5
424
425- IDE: Fix IME unintended message submission with Enter and Tab
426- IDE: Add "Open in Terminal" link in login screen
427- Fix unhandled OAuth expiration 401 API errors
428- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
429
430## 2.0.1
431
432- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
433- Various bug fixes and presentation improvements
434
435## 2.0.0
436
437- New native VS Code extension
438- Fresh coat of paint throughout the whole app
439- /rewind a conversation to undo code changes
440- /usage command to see plan limits
441- Tab to toggle thinking (sticky across sessions)
442- Ctrl-R to search history
443- Unshipped claude config command
444- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
445- SDK: The Claude Code SDK is now the Claude Agent SDK
446- Add subagents dynamically with `--agents` flag
447
448## 1.0.126
449
450- Enable /context command for Bedrock and Vertex
451- Add mTLS support for HTTP-based OpenTelemetry exporters
452
453## 1.0.124
454
455- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
456- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
457- No longer inform Claude of the list of allowed tools when permission is denied
458- Fixed security vulnerability in Bash tool permission checks
459- Improved VSCode extension performance for large files
460
461## 1.0.123
462
463- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
464- Fixed thinking mode triggering on negation phrases like "don't think"
465- Fixed rendering performance degradation during token streaming
466- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
467- Enhanced BashTool environment snapshot logging
468- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
469- Migrated --debug logging to a file, to enable easy tailing & filtering
470
471## 1.0.120
472
473- Fix input lag during typing, especially noticeable with large prompts
474- Improved VSCode extension command registry and sessions dialog user experience
475- Enhanced sessions dialog responsiveness and visual feedback
476- Fixed IDE compatibility issue by removing worktree support check
477- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
478
479## 1.0.119
480
481- Fix Windows issue where process visually freezes on entering interactive mode
482- Support dynamic headers for MCP servers via headersHelper configuration
483- Fix thinking mode not working in headless sessions
484- Fix slash commands now properly update allowed tools instead of replacing them
485
486## 1.0.117
487
488- Add Ctrl-R history search to recall previous commands like bash/zsh
489- Fix input lag while typing, especially on Windows
490- Add sed command to auto-allowed commands in acceptEdits mode
491- Fix Windows PATH comparison to be case-insensitive for drive letters
492- Add permissions management hint to /add-dir output
493
494## 1.0.115
495
496- Improve thinking mode display with enhanced visual effects
497- Type /t to temporarily disable thinking mode in your prompt
498- Improve path validation for glob and grep tools
499- Show condensed output for post-tool hooks to reduce visual clutter
500- Fix visual feedback when loading state completes
501- Improve UI consistency for permission request dialogs
502
503## 1.0.113
504
505- Deprecated piped input in interactive mode
506- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
507
508## 1.0.112
509
510- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
511- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
512- Hooks: Added systemMessage support for SessionEnd hooks
513- Added `spinnerTipsEnabled` setting to disable spinner tips
514- IDE: Various improvements and bug fixes
515
516## 1.0.111
517
518- /model now validates provided model names
519- Fixed Bash tool crashes caused by malformed shell syntax parsing
520
521## 1.0.110
522
523- /terminal-setup command now supports WezTerm
524- MCP: OAuth tokens now proactively refresh before expiration
525- Fixed reliability issues with background Bash processes
526
527## 1.0.109
528
529- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
530
531## 1.0.106
532
533- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
534
535## 1.0.97
536
537- Settings: /doctor now validates permission rule syntax and suggests corrections
538
539## 1.0.94
540
541- Vertex: add support for global endpoints for supported models
542- /memory command now allows direct editing of all imported memory files
543- SDK: Add custom tools as callbacks
544- Added /todos command to list current todo items
545
546## 1.0.93
547
548- Windows: Add alt + v shortcut for pasting images from clipboard
549- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
550
551## 1.0.90
552
553- Settings file changes take effect immediately - no restart required
554
555## 1.0.88
556
557- Fixed issue causing "OAuth authentication is currently not supported"
558- Status line input now includes `exceeds_200k_tokens`
559- Fixed incorrect usage tracking in /cost.
560- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
561- Bedrock: Updated default Sonnet model to Sonnet 4
562
563## 1.0.86
564
565- Added /context to help users self-serve debug context issues
566- SDK: Added UUID support for all SDK messages
567- SDK: Added `--replay-user-messages` to replay user messages back to stdout
568
569## 1.0.85
570
571- Status line input now includes session cost info
572- Hooks: Introduced SessionEnd hook
573
574## 1.0.84
575
576- Fix tool_use/tool_result id mismatch error when network is unstable
577- Fix Claude sometimes ignoring real-time steering when wrapping up a task
578- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
579- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
580
581## 1.0.83
582
583- @-mention: Support files with spaces in path
584- New shimmering spinner
585
586## 1.0.82
587
588- SDK: Add request cancellation support
589- SDK: New additionalDirectories option to search custom paths, improved slash command processing
590- Settings: Validation prevents invalid fields in .claude/settings.json files
591- MCP: Improve tool name consistency
592- Bash: Fix crash when Claude tries to automatically read large files
593
594## 1.0.81
595
596- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
597- Agents: Fix custom agent loading when agent files are unparsable
598
599## 1.0.80
600
601- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
602
603## 1.0.77
604
605- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
606- SDK: Add session support and permission denial tracking
607- Fix token limit errors in conversation summarization
608- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
609
610## 1.0.73
611
612- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
613- MCP: Press Esc to cancel OAuth authentication flows
614- Bash: Improved command validation and reduced false security warnings
615- UI: Enhanced spinner animations and status line visual hierarchy
616- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
617
618## 1.0.72
619
620- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
621
622## 1.0.71
623
624- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
625- Customizable status line: add your terminal prompt to Claude Code with /statusline
626
627## 1.0.70
628
629- Performance: Optimized message rendering for better performance with large contexts
630- Windows: Fixed native file search, ripgrep, and subagent functionality
631- Added support for @-mentions in slash command arguments
632
633## 1.0.69
634
635- Upgraded Opus to version 4.1
636
637## 1.0.68
638
639- Fix incorrect model names being used for certain commands like `/pr-comments`
640- 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.
641- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
642- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
643- SDK: Added canUseTool callback support for tool confirmation
644- Added `disableAllHooks` setting
645- Improved file suggestions performance in large repos
646
647## 1.0.65
648
649- IDE: Fixed connection stability issues and error handling for diagnostics
650- Windows: Fixed shell environment setup for users without .bashrc files
651
652## 1.0.64
653
654- Agents: Added model customization support - you can now specify which model an agent should use
655- Agents: Fixed unintended access to the recursive agent tool
656- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
657- SDK: Fixed user input tracking across multi-turn conversations
658- Added hidden files to file search and @-mention suggestions
659
660## 1.0.63
661
662- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
663
664## 1.0.62
665
666- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
667- Hooks: Added SessionStart hook for new session initialization
668- /add-dir command now supports typeahead for directory paths
669- Improved network connectivity check reliability
670
671## 1.0.61
672
673- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
674- Settings: Added `--settings` flag to load settings from a JSON file
675- Settings: Fixed resolution of settings files paths that are symlinks
676- OTEL: Fixed reporting of wrong organization after authentication changes
677- Slash commands: Fixed permissions checking for allowed-tools with Bash
678- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
679- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
680- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
681
682## 1.0.60
683
684- You can now create custom subagents for specialized tasks! Run /agents to get started
685
686## 1.0.59
687
688- SDK: Added tool confirmation support with canUseTool callback
689- SDK: Allow specifying env for spawned process
690- Hooks: Exposed PermissionDecision to hooks (including "ask")
691- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
692- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
693
694## 1.0.58
695
696- Added support for reading PDFs
697- MCP: Improved server health status display in 'claude mcp list'
698- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
699
700## 1.0.57
701
702- Added support for specifying a model in slash commands
703- Improved permission messages to help Claude understand allowed tools
704- Fix: Remove trailing newlines from bash output in terminal wrapping
705
706## 1.0.56
707
708- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
709- Fixes for WSL IDE detection
710- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
711
712## 1.0.55
713
714- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
715- Windows: fixed Ctrl+Z crash
716- SDK: Added ability to capture error logging
717- Add --system-prompt-file option to override system prompt in print mode
718
719## 1.0.54
720
721- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
722- Custom slash commands: Added argument-hint to frontmatter
723- Windows: OAuth uses port 45454 and properly constructs browser URL
724- Windows: mode switching now uses alt + m, and plan mode renders properly
725- Shell: Switch to in-memory shell snapshot to fix file-related errors
726
727## 1.0.53
728
729- Updated @-mention file truncation from 100 lines to 2000 lines
730- 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).
731
732## 1.0.52
733
734- Added support for MCP server instructions
735
736## 1.0.51
737
738- Added support for native Windows (requires Git for Windows)
739- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
740- Settings: /doctor can now help you identify and fix invalid setting files
741- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
742- Increased auto-compact warning threshold from 60% to 80%
743- Fixed an issue with handling user directories with spaces for shell snapshots
744- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
745- Custom slash commands: Fixed user-level commands in subdirectories
746- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
747
748## 1.0.48
749
750- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
751- Added progress messages to Bash tool based on the last 5 lines of command output
752- Added expanding variables support for MCP server configuration
753- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
754- Improved IDE extension path handling when Claude Code runs in WSL
755- Hooks: Added a PreCompact hook
756- Vim mode: Added c, f/F, t/T
757
758## 1.0.45
759
760- Redesigned Search (Grep) tool with new tool input parameters and features
761- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
762- Fixed config file corruption issue by enforcing atomic writes
763- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
764- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
765- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
766
767## 1.0.44
768
769- New /export command lets you quickly export a conversation for sharing
770- MCP: resource_link tool results are now supported
771- MCP: tool annotations and tool titles now display in /mcp view
772- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
773
774## 1.0.43
775
776- Fixed a bug where the theme selector was saving excessively
777- Hooks: Added EPIPE system error handling
778
779## 1.0.42
780
781- Added tilde (`~`) expansion support to `/add-dir` command
782
783## 1.0.41
784
785- Hooks: Split Stop hook triggering into Stop and SubagentStop
786- Hooks: Enabled optional timeout configuration for each command
787- Hooks: Added "hook_event_name" to hook input
788- Fixed a bug where MCP tools would display twice in tool list
789- New tool parameters JSON for Bash tool in `tool_decision` event
790
791## 1.0.40
792
793- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
794
795## 1.0.39
796
797- New Active Time metric in OpenTelemetry logging
798
799## 1.0.38
800
801- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
802
803## 1.0.37
804
805- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
806
807## 1.0.36
808
809- Web search now takes today's date into context
810- Fixed a bug where stdio MCP servers were not terminating properly on exit
811
812## 1.0.35
813
814- Added support for MCP OAuth Authorization Server discovery
815
816## 1.0.34
817
818- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
819
820## 1.0.33
821
822- Improved logging functionality with session ID support
823- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
824- Improvements to plan mode
825
826## 1.0.32
827
828- Updated loopback config for litellm
829- Added forceLoginMethod setting to bypass login selection screen
830
831## 1.0.31
832
833- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
834
835## 1.0.30
836
837- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
838- Improved file path autocomplete with filename matching
839- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
840- Enhanced jq regex support for complex filters with pipes and select
841
842## 1.0.29
843
844- Improved CJK character support in cursor navigation and rendering
845
846## 1.0.28
847
848- Slash commands: Fix selector display during history navigation
849- Resizes images before upload to prevent API size limit errors
850- Added XDG_CONFIG_HOME support to configuration directory
851- Performance optimizations for memory usage
852- New attributes (terminal.type, language) in OpenTelemetry logging
853
854## 1.0.27
855
856- Streamable HTTP MCP servers are now supported
857- Remote MCP servers (SSE and HTTP) now support OAuth
858- MCP resources can now be @-mentioned
859- /resume slash command to switch conversations within Claude Code
860
861## 1.0.25
862
863- Slash commands: moved "project" and "user" prefixes to descriptions
864- Slash commands: improved reliability for command discovery
865- Improved support for Ghostty
866- Improved web search reliability
867
868## 1.0.24
869
870- Improved /mcp output
871- Fixed a bug where settings arrays got overwritten instead of merged
872
873## 1.0.23
874
875- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
876- Released Python SDK: pip install claude-code-sdk to get started
877
878## 1.0.22
879
880- SDK: Renamed `total_cost` to `total_cost_usd`
881
882## 1.0.21
883
884- Improved editing of files with tab-based indentation
885- Fix for tool_use without matching tool_result errors
886- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
887
888## 1.0.18
889
890- Added --add-dir CLI argument for specifying additional working directories
891- Added streaming input support without require -p flag
892- Improved startup performance and session storage performance
893- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
894- Added detailed MCP server tools display (/mcp)
895- MCP authentication and permission improvements
896- Added auto-reconnection for MCP SSE connections on disconnect
897- Fixed issue where pasted content was lost when dialogs appeared
898
899## 1.0.17
900
901- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
902- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
903- MCP server list UI improvements
904- Update Claude Code process title to display "claude" instead of "node"
905
906## 1.0.11
907
908- Claude Code can now also be used with a Claude Pro subscription
909- Added /upgrade for smoother switching to Claude Max plans
910- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
911- Improved shell configuration error handling
912- Improved todo list handling during compaction
913
914## 1.0.10
915
916- Added markdown table support
917- Improved streaming performance
918
919## 1.0.8
920
921- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
922- Increased default otel interval from 1s -> 5s
923- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
924- Fixed a regression where search tools unnecessarily asked for permissions
925- Added support for triggering thinking non-English languages
926- Improved compacting UI
927
928## 1.0.7
929
930- Renamed /allowed-tools -> /permissions
931- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
932- Deprecated claude config commands in favor of editing settings.json
933- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
934- Improved error handling for /install-github-app
935- Bugfixes, UI polish, and tool reliability improvements
936
937## 1.0.6
938
939- Improved edit reliability for tab-indented files
940- Respect CLAUDE_CONFIG_DIR everywhere
941- Reduced unnecessary tool permission prompts
942- Added support for symlinks in @file typeahead
943- Bugfixes, UI polish, and tool reliability improvements
944
945## 1.0.4
946
947- Fixed a bug where MCP tool errors weren't being parsed correctly
948
949## 1.0.1
950
951- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
952- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
953- Updated documentation links and OAuth process descriptions
954
955## 1.0.0
956
957- Claude Code is now generally available
958- Introducing Sonnet 4 and Opus 4 models
959
960## 0.2.125
961
962- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
963- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
964
965## 0.2.117
966
967- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
968- Introduced settings.cleanupPeriodDays
969- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
970- Introduced --debug mode
971
972## 0.2.108
973
974- You can now send messages to Claude while it works to steer Claude in real-time
975- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
976- Fixed a bug where thinking was not working in -p mode
977- Fixed a regression in /cost reporting
978- Deprecated MCP wizard interface in favor of other MCP commands
979- Lots of other bugfixes and improvements
980
981## 0.2.107
982
983- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
984
985## 0.2.106
986
987- MCP SSE server configs can now specify custom headers
988- Fixed a bug where MCP permission prompt didn't always show correctly
989
990## 0.2.105
991
992- Claude can now search the web
993- Moved system & account status to /status
994- Added word movement keybindings for Vim
995- Improved latency for startup, todo tool, and file edits
996
997## 0.2.102
998
999- Improved thinking triggering reliability
1000- Improved @mention reliability for images and folders
1001- You can now paste multiple large chunks into one prompt
1002
1003## 0.2.100
1004
1005- Fixed a crash caused by a stack overflow error
1006- Made db storage optional; missing db support disables --continue and --resume
1007
1008## 0.2.98
1009
1010- Fixed an issue where auto-compact was running twice
1011
1012## 0.2.96
1013
1014- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1015
1016## 0.2.93
1017
1018- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1019- Claude now has access to a Todo list that helps it stay on track and be more organized
1020
1021## 0.2.82
1022
1023- Added support for --disallowedTools
1024- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1025
1026## 0.2.75
1027
1028- Hit Enter to queue up additional messages while Claude is working
1029- Drag in or copy/paste image files directly into the prompt
1030- @-mention files to directly add them to context
1031- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1032- Improved performance for filename auto-complete
1033
1034## 0.2.74
1035
1036- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1037- Task tool can now perform writes and run bash commands
1038
1039## 0.2.72
1040
1041- Updated spinner to indicate tokens loaded and tool usage
1042
1043## 0.2.70
1044
1045- Network commands like curl are now available for Claude to use
1046- Claude can now run multiple web queries in parallel
1047- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1048
1049## 0.2.69
1050
1051- Fixed UI glitches with improved Select component behavior
1052- Enhanced terminal output display with better text truncation logic
1053
1054## 0.2.67
1055
1056- Shared project permission rules can be saved in .claude/settings.json
1057
1058## 0.2.66
1059
1060- Print mode (-p) now supports streaming output via --output-format=stream-json
1061- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1062
1063## 0.2.63
1064
1065- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1066
1067## 0.2.61
1068
1069- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1070- Enhanced image detection for more reliable clipboard paste functionality
1071- Fixed an issue where ESC key could crash the conversation history selector
1072
1073## 0.2.59
1074
1075- Copy+paste images directly into your prompt
1076- Improved progress indicators for bash and fetch tools
1077- Bugfixes for non-interactive mode (-p)
1078
1079## 0.2.54
1080
1081- Quickly add to Memory by starting your message with '#'
1082- Press ctrl+r to see full output for long tool results
1083- Added support for MCP SSE transport
1084
1085## 0.2.53
1086
1087- New web fetch tool lets Claude view URLs that you paste in
1088- Fixed a bug with JPEG detection
1089
1090## 0.2.50
1091
1092- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1093
1094## 0.2.49
1095
1096- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1097
1098## 0.2.47
1099
1100- Press Tab to auto-complete file and folder names
1101- Press Shift + Tab to toggle auto-accept for file edits
1102- Automatic conversation compaction for infinite conversation length (toggle with /config)
1103
1104## 0.2.44
1105
1106- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1107
1108## 0.2.41
1109
1110- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1111- MCP server startup no longer blocks the app from starting up
1112
1113## 0.2.37
1114
1115- New /release-notes command lets you view release notes at any time
1116- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1117
1118## 0.2.36
1119
1120- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1121- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1122
1123## 0.2.34
1124
1125- Vim bindings for text input - enable with /vim or /config
1126
1127## 0.2.32
1128
1129- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1130- Fix for some PersistentShell issues
1131
1132## 0.2.31
1133
1134- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1135- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1136
1137## 0.2.30
1138
1139- Added ANSI color theme for better terminal compatibility
1140- Fixed issue where slash command arguments weren't being sent properly
1141- (Mac-only) API keys are now stored in macOS Keychain
1142
1143## 0.2.26
1144
1145- New /approved-tools command for managing tool permissions
1146- Word-level diff display for improved code readability
1147- Fuzzy matching for slash commands
1148
1149## 0.2.21
1150
1151- Fuzzy matching for /commands