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 5211559 · on Dec 19, 2025 History
1,162 lines · 48.1 KBPreviewCodeRaw
1# Changelog
2
3## 2.0.73
4
5- Added clickable `[Image #N]` links that open attached images in the default viewer
6- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
7- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
8- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
9- Fixed slow input history cycling and race condition that could overwrite text after message submission
10- Improved `/theme` command to open theme picker directly
11- Improved theme picker UI
12- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
13- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
14
15## 2.0.72
16
17- 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
18- Reduced terminal flickering
19- Added scannable QR code to mobile app tip for quick app downloads
20- Added loading indicator when resuming conversations for better feedback
21- Fixed `/context` command not respecting custom system prompts in non-interactive mode
22- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
23- Improved @ mention file suggestion speed (~3x faster in git repositories)
24- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
25- Improved settings validation errors to be more prominent
26- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
27
28## 2.0.71
29
30- Added /config toggle to enable/disable prompt suggestions
31- Added `/settings` as an alias for the `/config` command
32- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
33- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
34- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
35- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
36- New syntax highlighting engine for native build
37
38## 2.0.70
39
40- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
41- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
42- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
43- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
44- Fixed input being cleared when processing queued commands while the user was typing
45- Fixed prompt suggestions replacing typed input when pressing Tab
46- Fixed diff view not updating when terminal is resized
47- Improved memory usage by 3x for large conversations
48- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
49- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
50- Fix thinking mode toggle in /config not persisting correctly
51- Improve UI for file creation permission dialog
52
53## 2.0.69
54
55- Minor bugfixes
56
57## 2.0.68
58
59- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
60- Fixed a bug where disallowed MCP tools were visible to the model
61- Fixed an issue where steering messages could be lost while a subagent is working
62- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
63- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
64- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
65
66## 2.0.67
67
68- Thinking mode is now enabled by default for Opus 4.5
69- Thinking mode configuration has moved to /config
70- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
71- Show reason why autoupdater is disabled in `/doctor`
72- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
73- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
74- Fixed scroll position resetting after deleting a permission rule in `/permissions`
75- 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
76- Fixed `claude install --force` not bypassing stale lock files
77- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
78- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
79
80## 2.0.65
81
82- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
83- Added context window information to status line input
84- Added `fileSuggestion` setting for custom `@` file search commands
85- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
86- Fixed prompt not being saved to history when aborting a query with Escape
87- Fixed Read tool image handling to identify format from bytes instead of file extension
88
89## 2.0.64
90
91- Made auto-compacting instant
92- Agents and bash commands can run asynchronously and send messages to wake up the main agent
93- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
94- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
95- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
96- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
97- Fixed auto-loading .env when using native installer
98- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
99- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
100- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
101- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
102- Bedrock: Improve efficiency of token counting
103- Bedrock: Add support for `aws login` AWS Management Console credentials
104- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
105
106## 2.0.62
107
108- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
109- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
110- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
111- Fixed slash command selection not working when multiple commands share the same name
112- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
113- Fixed running versions getting removed because lock file incorrectly going stale
114- Fixed IDE diff tab not closing when rejecting file changes
115
116## 2.0.61
117
118- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
119
120## 2.0.60
121
122- Added background agent support. Agents run in the background while you work
123- Added --disable-slash-commands CLI flag to disable all slash commands
124- Added model name to "Co-Authored-By" commit messages
125- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
126- Updated Fetch to skip summarization for pre-approved websites
127- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
128
129## 2.0.59
130
131- Added --agent CLI flag to override the agent setting for the current session
132- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
133- VS Code: Fixed .claude.json config file being read from incorrect location
134
135## 2.0.58
136
137- Pro users now have access to Opus 4.5 as part of their subscription!
138- Fixed timer duration showing "11m 60s" instead of "12m 0s"
139- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
140
141## 2.0.57
142
143- Added feedback input when rejecting plans, allowing users to tell Claude what to change
144- VSCode: Added streaming message support for real-time response display
145
146## 2.0.56
147
148- Added setting to enable/disable terminal progress bar (OSC 9;4)
149- 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.
150
151## 2.0.55
152
153- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
154- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
155- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
156- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
157
158## 2.0.54
159
160- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
161- Fix issue with excessive iTerm notifications
162
163## 2.0.52
164
165- Fixed duplicate message display when starting Claude with a command line argument
166- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
167- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
168- Permit some uses of `$!` in bash commands
169
170## 2.0.51
171
172- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
173- Introducing Claude Code for Desktop: https://claude.com/download
174- 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
175- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
176- Plan Mode now builds more precise plans and executes more thoroughly
177- Usage limit notifications now easier to understand
178- Switched `/usage` back to "% used"
179- Fixed handling of thinking errors
180- Fixed performance regression
181
182## 2.0.50
183
184- Fixed bug preventing calling MCP tools that have nested references in their input schemas
185- Silenced a noisy but harmless error during upgrades
186- Improved ultrathink text display
187- Improved clarity of 5-hour session limit warning message
188
189## 2.0.49
190
191- Added readline-style ctrl-y for pasting deleted text
192- Improved clarity of usage limit warning message
193- Fixed handling of subagent permissions
194
195## 2.0.47
196
197- Improved error messages and validation for `claude --teleport`
198- Improved error handling in `/usage`
199- Fixed race condition with history entry not getting logged at exit
200- Fixed Vertex AI configuration not being applied from `settings.json`
201
202## 2.0.46
203
204- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
205
206## 2.0.45
207
208- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
209- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
210- Send background tasks to Claude Code on the web by starting a message with `&`
211
212## 2.0.43
213
214- Added `permissionMode` field for custom agents
215- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
216- Added skills frontmatter field to declare skills to auto-load for subagents
217- Added the `SubagentStart` hook event
218- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
219- Fixed duplicate rendering of some messages in the UI
220- Fixed some visual flickers
221- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
222
223## 2.0.42
224
225- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
226
227## 2.0.41
228
229- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
230- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
231- Fixed incorrect labeling of user settings vs project settings in command descriptions
232- Fixed crash when plugin command hooks timeout during execution
233- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
234- Fixed broken security documentation links in trust dialogs and onboarding
235- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
236- ctrl-r history search landing on a slash command no longer cancels the search
237- SDK: Support custom timeouts for hooks
238- Allow more safe git commands to run without approval
239- Plugins: Added support for sharing and installing output styles
240- Teleporting a session from web will automatically set the upstream branch
241
242## 2.0.37
243
244- Fixed how idleness is computed for notifications
245- Hooks: Added matcher values for Notification hook events
246- Output Styles: Added `keep-coding-instructions` option to frontmatter
247
248## 2.0.36
249
250- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
251- Fixed queued messages being incorrectly executed as bash commands
252- Fixed input being lost when typing while a queued message is processed
253
254## 2.0.35
255
256- Improve fuzzy search results when searching commands
257- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
258- 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
259- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
260- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
261
262## 2.0.34
263
264- VSCode Extension: Added setting to configure the initial permission mode for new conversations
265- Improved file path suggestion performance with native Rust-based fuzzy finder
266- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
267- Fixed memory crash when reading or writing large files (especially base64-encoded images)
268
269## 2.0.33
270
271- Native binary installs now launch quicker.
272- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
273- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
274
275## 2.0.32
276
277- Un-deprecate output styles based on community feedback
278- Added `companyAnnouncements` setting for displaying announcements on startup
279- Fixed hook progress messages not updating correctly during PostToolUse hook execution
280
281## 2.0.31
282
283- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
284- Vertex: add support for Web Search on supported models
285- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
286- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
287- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
288- Fixed plugin uninstall not removing plugins
289
290## 2.0.30
291
292- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
293- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
294- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
295- Added prompt-based stop hooks
296- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
297- Enabled SSE MCP servers on native build
298- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
299- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
300- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
301- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
302- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
303- Fixed bug that saved session permissions to local settings
304- Fixed MCP tools not being available to sub-agents
305- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
306- Fixed delay when navigating through typeahead suggestions with arrow keys
307- VSCode: Restored selection indicator in input footer showing current file or code selection status
308
309## 2.0.28
310
311- Plan mode: introduced new Plan subagent
312- Subagents: claude can now choose to resume subagents
313- Subagents: claude can dynamically choose the model used by its subagents
314- SDK: added --max-budget-usd flag
315- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
316- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
317- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
318- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
319- Various other bug fixes
320
321## 2.0.27
322
323- New UI for permission prompts
324- Added current branch filtering and search to session resume screen for easier navigation
325- Fixed directory @-mention causing "No assistant message found" error
326- VSCode Extension: Add config setting to include .gitignored files in file searches
327- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
328
329## 2.0.25
330
331- 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
332
333## 2.0.24
334
335- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
336- Claude Code Web: Support for Web -> CLI teleport
337- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
338- Bedrock: Display awsAuthRefresh output when auth is required
339
340## 2.0.22
341
342- Fixed content layout shift when scrolling through slash commands
343- IDE: Add toggle to enable/disable thinking.
344- Fix bug causing duplicate permission prompts with parallel tool calls
345- Add support for enterprise managed MCP allowlist and denylist
346
347## 2.0.21
348
349- Support MCP `structuredContent` field in tool responses
350- Added an interactive question tool
351- Claude will now ask you questions more often in plan mode
352- Added Haiku 4.5 as a model option for Pro users
353- Fixed an issue where queued commands don't have access to previous messages' output
354
355## 2.0.20
356
357- Added support for Claude Skills
358
359## 2.0.19
360
361- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
362- Fixed a bug where Haiku was unnecessarily called in print mode
363
364## 2.0.17
365
366- Added Haiku 4.5 to model selector!
367- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
368- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
369- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
370- OTEL: support HTTP_PROXY and HTTPS_PROXY
371- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
372
373## 2.0.15
374
375- Fixed bug with resuming where previously created files needed to be read again before writing
376- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
377
378## 2.0.14
379
380- Fix @-mentioning MCP servers to toggle them on/off
381- Improve permission checks for bash with inline env vars
382- Fix ultrathink + thinking toggle
383- Reduce unnecessary logins
384- Document --system-prompt
385- Several improvements to rendering
386- Plugins UI polish
387
388## 2.0.13
389
390- Fixed `/plugin` not working on native build
391
392## 2.0.12
393
394- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
395- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
396- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
397- `/plugin validate` command for validating plugin structure and configuration
398- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
399- Plugin documentation available at https://code.claude.com/docs/en/plugins
400- Comprehensive error messages and diagnostics via `/doctor` command
401- Avoid flickering in `/model` selector
402- Improvements to `/help`
403- Avoid mentioning hooks in `/resume` summaries
404- Changes to the "verbose" setting in `/config` now persist across sessions
405
406## 2.0.11
407
408- Reduced system prompt size by 1.4k tokens
409- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
410- Fixed Opus fallback rate limit errors appearing incorrectly
411- Fixed /add-dir command selecting wrong default tab
412
413## 2.0.10
414
415- Rewrote terminal renderer for buttery smooth UI
416- Enable/disable MCP servers by @mentioning, or in /mcp
417- Added tab completion for shell commands in bash mode
418- PreToolUse hooks can now modify tool inputs
419- Press Ctrl-G to edit your prompt in your system's configured text editor
420- Fixes for bash permission checks with environment variables in the command
421
422## 2.0.9
423
424- Fix regression where bash backgrounding stopped working
425
426## 2.0.8
427
428- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
429- IDE: Add drag-and-drop support for files and folders in chat
430- /context: Fix counting for thinking blocks
431- Improve message rendering for users with light themes on dark terminals
432- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
433
434## 2.0.5
435
436- IDE: Fix IME unintended message submission with Enter and Tab
437- IDE: Add "Open in Terminal" link in login screen
438- Fix unhandled OAuth expiration 401 API errors
439- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
440
441## 2.0.1
442
443- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
444- Various bug fixes and presentation improvements
445
446## 2.0.0
447
448- New native VS Code extension
449- Fresh coat of paint throughout the whole app
450- /rewind a conversation to undo code changes
451- /usage command to see plan limits
452- Tab to toggle thinking (sticky across sessions)
453- Ctrl-R to search history
454- Unshipped claude config command
455- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
456- SDK: The Claude Code SDK is now the Claude Agent SDK
457- Add subagents dynamically with `--agents` flag
458
459## 1.0.126
460
461- Enable /context command for Bedrock and Vertex
462- Add mTLS support for HTTP-based OpenTelemetry exporters
463
464## 1.0.124
465
466- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
467- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
468- No longer inform Claude of the list of allowed tools when permission is denied
469- Fixed security vulnerability in Bash tool permission checks
470- Improved VSCode extension performance for large files
471
472## 1.0.123
473
474- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
475- Fixed thinking mode triggering on negation phrases like "don't think"
476- Fixed rendering performance degradation during token streaming
477- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
478- Enhanced BashTool environment snapshot logging
479- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
480- Migrated --debug logging to a file, to enable easy tailing & filtering
481
482## 1.0.120
483
484- Fix input lag during typing, especially noticeable with large prompts
485- Improved VSCode extension command registry and sessions dialog user experience
486- Enhanced sessions dialog responsiveness and visual feedback
487- Fixed IDE compatibility issue by removing worktree support check
488- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
489
490## 1.0.119
491
492- Fix Windows issue where process visually freezes on entering interactive mode
493- Support dynamic headers for MCP servers via headersHelper configuration
494- Fix thinking mode not working in headless sessions
495- Fix slash commands now properly update allowed tools instead of replacing them
496
497## 1.0.117
498
499- Add Ctrl-R history search to recall previous commands like bash/zsh
500- Fix input lag while typing, especially on Windows
501- Add sed command to auto-allowed commands in acceptEdits mode
502- Fix Windows PATH comparison to be case-insensitive for drive letters
503- Add permissions management hint to /add-dir output
504
505## 1.0.115
506
507- Improve thinking mode display with enhanced visual effects
508- Type /t to temporarily disable thinking mode in your prompt
509- Improve path validation for glob and grep tools
510- Show condensed output for post-tool hooks to reduce visual clutter
511- Fix visual feedback when loading state completes
512- Improve UI consistency for permission request dialogs
513
514## 1.0.113
515
516- Deprecated piped input in interactive mode
517- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
518
519## 1.0.112
520
521- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
522- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
523- Hooks: Added systemMessage support for SessionEnd hooks
524- Added `spinnerTipsEnabled` setting to disable spinner tips
525- IDE: Various improvements and bug fixes
526
527## 1.0.111
528
529- /model now validates provided model names
530- Fixed Bash tool crashes caused by malformed shell syntax parsing
531
532## 1.0.110
533
534- /terminal-setup command now supports WezTerm
535- MCP: OAuth tokens now proactively refresh before expiration
536- Fixed reliability issues with background Bash processes
537
538## 1.0.109
539
540- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
541
542## 1.0.106
543
544- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
545
546## 1.0.97
547
548- Settings: /doctor now validates permission rule syntax and suggests corrections
549
550## 1.0.94
551
552- Vertex: add support for global endpoints for supported models
553- /memory command now allows direct editing of all imported memory files
554- SDK: Add custom tools as callbacks
555- Added /todos command to list current todo items
556
557## 1.0.93
558
559- Windows: Add alt + v shortcut for pasting images from clipboard
560- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
561
562## 1.0.90
563
564- Settings file changes take effect immediately - no restart required
565
566## 1.0.88
567
568- Fixed issue causing "OAuth authentication is currently not supported"
569- Status line input now includes `exceeds_200k_tokens`
570- Fixed incorrect usage tracking in /cost.
571- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
572- Bedrock: Updated default Sonnet model to Sonnet 4
573
574## 1.0.86
575
576- Added /context to help users self-serve debug context issues
577- SDK: Added UUID support for all SDK messages
578- SDK: Added `--replay-user-messages` to replay user messages back to stdout
579
580## 1.0.85
581
582- Status line input now includes session cost info
583- Hooks: Introduced SessionEnd hook
584
585## 1.0.84
586
587- Fix tool_use/tool_result id mismatch error when network is unstable
588- Fix Claude sometimes ignoring real-time steering when wrapping up a task
589- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
590- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
591
592## 1.0.83
593
594- @-mention: Support files with spaces in path
595- New shimmering spinner
596
597## 1.0.82
598
599- SDK: Add request cancellation support
600- SDK: New additionalDirectories option to search custom paths, improved slash command processing
601- Settings: Validation prevents invalid fields in .claude/settings.json files
602- MCP: Improve tool name consistency
603- Bash: Fix crash when Claude tries to automatically read large files
604
605## 1.0.81
606
607- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
608- Agents: Fix custom agent loading when agent files are unparsable
609
610## 1.0.80
611
612- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
613
614## 1.0.77
615
616- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
617- SDK: Add session support and permission denial tracking
618- Fix token limit errors in conversation summarization
619- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
620
621## 1.0.73
622
623- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
624- MCP: Press Esc to cancel OAuth authentication flows
625- Bash: Improved command validation and reduced false security warnings
626- UI: Enhanced spinner animations and status line visual hierarchy
627- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
628
629## 1.0.72
630
631- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
632
633## 1.0.71
634
635- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
636- Customizable status line: add your terminal prompt to Claude Code with /statusline
637
638## 1.0.70
639
640- Performance: Optimized message rendering for better performance with large contexts
641- Windows: Fixed native file search, ripgrep, and subagent functionality
642- Added support for @-mentions in slash command arguments
643
644## 1.0.69
645
646- Upgraded Opus to version 4.1
647
648## 1.0.68
649
650- Fix incorrect model names being used for certain commands like `/pr-comments`
651- 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.
652- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
653- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
654- SDK: Added canUseTool callback support for tool confirmation
655- Added `disableAllHooks` setting
656- Improved file suggestions performance in large repos
657
658## 1.0.65
659
660- IDE: Fixed connection stability issues and error handling for diagnostics
661- Windows: Fixed shell environment setup for users without .bashrc files
662
663## 1.0.64
664
665- Agents: Added model customization support - you can now specify which model an agent should use
666- Agents: Fixed unintended access to the recursive agent tool
667- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
668- SDK: Fixed user input tracking across multi-turn conversations
669- Added hidden files to file search and @-mention suggestions
670
671## 1.0.63
672
673- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
674
675## 1.0.62
676
677- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
678- Hooks: Added SessionStart hook for new session initialization
679- /add-dir command now supports typeahead for directory paths
680- Improved network connectivity check reliability
681
682## 1.0.61
683
684- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
685- Settings: Added `--settings` flag to load settings from a JSON file
686- Settings: Fixed resolution of settings files paths that are symlinks
687- OTEL: Fixed reporting of wrong organization after authentication changes
688- Slash commands: Fixed permissions checking for allowed-tools with Bash
689- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
690- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
691- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
692
693## 1.0.60
694
695- You can now create custom subagents for specialized tasks! Run /agents to get started
696
697## 1.0.59
698
699- SDK: Added tool confirmation support with canUseTool callback
700- SDK: Allow specifying env for spawned process
701- Hooks: Exposed PermissionDecision to hooks (including "ask")
702- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
703- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
704
705## 1.0.58
706
707- Added support for reading PDFs
708- MCP: Improved server health status display in 'claude mcp list'
709- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
710
711## 1.0.57
712
713- Added support for specifying a model in slash commands
714- Improved permission messages to help Claude understand allowed tools
715- Fix: Remove trailing newlines from bash output in terminal wrapping
716
717## 1.0.56
718
719- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
720- Fixes for WSL IDE detection
721- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
722
723## 1.0.55
724
725- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
726- Windows: fixed Ctrl+Z crash
727- SDK: Added ability to capture error logging
728- Add --system-prompt-file option to override system prompt in print mode
729
730## 1.0.54
731
732- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
733- Custom slash commands: Added argument-hint to frontmatter
734- Windows: OAuth uses port 45454 and properly constructs browser URL
735- Windows: mode switching now uses alt + m, and plan mode renders properly
736- Shell: Switch to in-memory shell snapshot to fix file-related errors
737
738## 1.0.53
739
740- Updated @-mention file truncation from 100 lines to 2000 lines
741- 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).
742
743## 1.0.52
744
745- Added support for MCP server instructions
746
747## 1.0.51
748
749- Added support for native Windows (requires Git for Windows)
750- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
751- Settings: /doctor can now help you identify and fix invalid setting files
752- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
753- Increased auto-compact warning threshold from 60% to 80%
754- Fixed an issue with handling user directories with spaces for shell snapshots
755- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
756- Custom slash commands: Fixed user-level commands in subdirectories
757- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
758
759## 1.0.48
760
761- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
762- Added progress messages to Bash tool based on the last 5 lines of command output
763- Added expanding variables support for MCP server configuration
764- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
765- Improved IDE extension path handling when Claude Code runs in WSL
766- Hooks: Added a PreCompact hook
767- Vim mode: Added c, f/F, t/T
768
769## 1.0.45
770
771- Redesigned Search (Grep) tool with new tool input parameters and features
772- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
773- Fixed config file corruption issue by enforcing atomic writes
774- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
775- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
776- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
777
778## 1.0.44
779
780- New /export command lets you quickly export a conversation for sharing
781- MCP: resource_link tool results are now supported
782- MCP: tool annotations and tool titles now display in /mcp view
783- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
784
785## 1.0.43
786
787- Fixed a bug where the theme selector was saving excessively
788- Hooks: Added EPIPE system error handling
789
790## 1.0.42
791
792- Added tilde (`~`) expansion support to `/add-dir` command
793
794## 1.0.41
795
796- Hooks: Split Stop hook triggering into Stop and SubagentStop
797- Hooks: Enabled optional timeout configuration for each command
798- Hooks: Added "hook_event_name" to hook input
799- Fixed a bug where MCP tools would display twice in tool list
800- New tool parameters JSON for Bash tool in `tool_decision` event
801
802## 1.0.40
803
804- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
805
806## 1.0.39
807
808- New Active Time metric in OpenTelemetry logging
809
810## 1.0.38
811
812- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
813
814## 1.0.37
815
816- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
817
818## 1.0.36
819
820- Web search now takes today's date into context
821- Fixed a bug where stdio MCP servers were not terminating properly on exit
822
823## 1.0.35
824
825- Added support for MCP OAuth Authorization Server discovery
826
827## 1.0.34
828
829- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
830
831## 1.0.33
832
833- Improved logging functionality with session ID support
834- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
835- Improvements to plan mode
836
837## 1.0.32
838
839- Updated loopback config for litellm
840- Added forceLoginMethod setting to bypass login selection screen
841
842## 1.0.31
843
844- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
845
846## 1.0.30
847
848- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
849- Improved file path autocomplete with filename matching
850- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
851- Enhanced jq regex support for complex filters with pipes and select
852
853## 1.0.29
854
855- Improved CJK character support in cursor navigation and rendering
856
857## 1.0.28
858
859- Slash commands: Fix selector display during history navigation
860- Resizes images before upload to prevent API size limit errors
861- Added XDG_CONFIG_HOME support to configuration directory
862- Performance optimizations for memory usage
863- New attributes (terminal.type, language) in OpenTelemetry logging
864
865## 1.0.27
866
867- Streamable HTTP MCP servers are now supported
868- Remote MCP servers (SSE and HTTP) now support OAuth
869- MCP resources can now be @-mentioned
870- /resume slash command to switch conversations within Claude Code
871
872## 1.0.25
873
874- Slash commands: moved "project" and "user" prefixes to descriptions
875- Slash commands: improved reliability for command discovery
876- Improved support for Ghostty
877- Improved web search reliability
878
879## 1.0.24
880
881- Improved /mcp output
882- Fixed a bug where settings arrays got overwritten instead of merged
883
884## 1.0.23
885
886- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
887- Released Python SDK: pip install claude-code-sdk to get started
888
889## 1.0.22
890
891- SDK: Renamed `total_cost` to `total_cost_usd`
892
893## 1.0.21
894
895- Improved editing of files with tab-based indentation
896- Fix for tool_use without matching tool_result errors
897- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
898
899## 1.0.18
900
901- Added --add-dir CLI argument for specifying additional working directories
902- Added streaming input support without require -p flag
903- Improved startup performance and session storage performance
904- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
905- Added detailed MCP server tools display (/mcp)
906- MCP authentication and permission improvements
907- Added auto-reconnection for MCP SSE connections on disconnect
908- Fixed issue where pasted content was lost when dialogs appeared
909
910## 1.0.17
911
912- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
913- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
914- MCP server list UI improvements
915- Update Claude Code process title to display "claude" instead of "node"
916
917## 1.0.11
918
919- Claude Code can now also be used with a Claude Pro subscription
920- Added /upgrade for smoother switching to Claude Max plans
921- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
922- Improved shell configuration error handling
923- Improved todo list handling during compaction
924
925## 1.0.10
926
927- Added markdown table support
928- Improved streaming performance
929
930## 1.0.8
931
932- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
933- Increased default otel interval from 1s -> 5s
934- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
935- Fixed a regression where search tools unnecessarily asked for permissions
936- Added support for triggering thinking non-English languages
937- Improved compacting UI
938
939## 1.0.7
940
941- Renamed /allowed-tools -> /permissions
942- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
943- Deprecated claude config commands in favor of editing settings.json
944- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
945- Improved error handling for /install-github-app
946- Bugfixes, UI polish, and tool reliability improvements
947
948## 1.0.6
949
950- Improved edit reliability for tab-indented files
951- Respect CLAUDE_CONFIG_DIR everywhere
952- Reduced unnecessary tool permission prompts
953- Added support for symlinks in @file typeahead
954- Bugfixes, UI polish, and tool reliability improvements
955
956## 1.0.4
957
958- Fixed a bug where MCP tool errors weren't being parsed correctly
959
960## 1.0.1
961
962- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
963- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
964- Updated documentation links and OAuth process descriptions
965
966## 1.0.0
967
968- Claude Code is now generally available
969- Introducing Sonnet 4 and Opus 4 models
970
971## 0.2.125
972
973- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
974- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
975
976## 0.2.117
977
978- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
979- Introduced settings.cleanupPeriodDays
980- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
981- Introduced --debug mode
982
983## 0.2.108
984
985- You can now send messages to Claude while it works to steer Claude in real-time
986- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
987- Fixed a bug where thinking was not working in -p mode
988- Fixed a regression in /cost reporting
989- Deprecated MCP wizard interface in favor of other MCP commands
990- Lots of other bugfixes and improvements
991
992## 0.2.107
993
994- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
995
996## 0.2.106
997
998- MCP SSE server configs can now specify custom headers
999- Fixed a bug where MCP permission prompt didn't always show correctly
1000
1001## 0.2.105
1002
1003- Claude can now search the web
1004- Moved system & account status to /status
1005- Added word movement keybindings for Vim
1006- Improved latency for startup, todo tool, and file edits
1007
1008## 0.2.102
1009
1010- Improved thinking triggering reliability
1011- Improved @mention reliability for images and folders
1012- You can now paste multiple large chunks into one prompt
1013
1014## 0.2.100
1015
1016- Fixed a crash caused by a stack overflow error
1017- Made db storage optional; missing db support disables --continue and --resume
1018
1019## 0.2.98
1020
1021- Fixed an issue where auto-compact was running twice
1022
1023## 0.2.96
1024
1025- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1026
1027## 0.2.93
1028
1029- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1030- Claude now has access to a Todo list that helps it stay on track and be more organized
1031
1032## 0.2.82
1033
1034- Added support for --disallowedTools
1035- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1036
1037## 0.2.75
1038
1039- Hit Enter to queue up additional messages while Claude is working
1040- Drag in or copy/paste image files directly into the prompt
1041- @-mention files to directly add them to context
1042- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1043- Improved performance for filename auto-complete
1044
1045## 0.2.74
1046
1047- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1048- Task tool can now perform writes and run bash commands
1049
1050## 0.2.72
1051
1052- Updated spinner to indicate tokens loaded and tool usage
1053
1054## 0.2.70
1055
1056- Network commands like curl are now available for Claude to use
1057- Claude can now run multiple web queries in parallel
1058- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1059
1060## 0.2.69
1061
1062- Fixed UI glitches with improved Select component behavior
1063- Enhanced terminal output display with better text truncation logic
1064
1065## 0.2.67
1066
1067- Shared project permission rules can be saved in .claude/settings.json
1068
1069## 0.2.66
1070
1071- Print mode (-p) now supports streaming output via --output-format=stream-json
1072- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1073
1074## 0.2.63
1075
1076- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1077
1078## 0.2.61
1079
1080- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1081- Enhanced image detection for more reliable clipboard paste functionality
1082- Fixed an issue where ESC key could crash the conversation history selector
1083
1084## 0.2.59
1085
1086- Copy+paste images directly into your prompt
1087- Improved progress indicators for bash and fetch tools
1088- Bugfixes for non-interactive mode (-p)
1089
1090## 0.2.54
1091
1092- Quickly add to Memory by starting your message with '#'
1093- Press ctrl+r to see full output for long tool results
1094- Added support for MCP SSE transport
1095
1096## 0.2.53
1097
1098- New web fetch tool lets Claude view URLs that you paste in
1099- Fixed a bug with JPEG detection
1100
1101## 0.2.50
1102
1103- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1104
1105## 0.2.49
1106
1107- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1108
1109## 0.2.47
1110
1111- Press Tab to auto-complete file and folder names
1112- Press Shift + Tab to toggle auto-accept for file edits
1113- Automatic conversation compaction for infinite conversation length (toggle with /config)
1114
1115## 0.2.44
1116
1117- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1118
1119## 0.2.41
1120
1121- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1122- MCP server startup no longer blocks the app from starting up
1123
1124## 0.2.37
1125
1126- New /release-notes command lets you view release notes at any time
1127- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1128
1129## 0.2.36
1130
1131- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1132- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1133
1134## 0.2.34
1135
1136- Vim bindings for text input - enable with /vim or /config
1137
1138## 0.2.32
1139
1140- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1141- Fix for some PersistentShell issues
1142
1143## 0.2.31
1144
1145- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1146- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1147
1148## 0.2.30
1149
1150- Added ANSI color theme for better terminal compatibility
1151- Fixed issue where slash command arguments weren't being sent properly
1152- (Mac-only) API keys are now stored in macOS Keychain
1153
1154## 0.2.26
1155
1156- New /approved-tools command for managing tool permissions
1157- Word-level diff display for improved code readability
1158- Fuzzy matching for slash commands
1159
1160## 0.2.21
1161
1162- Fuzzy matching for /commands