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 f860f67 · on Jan 13, 2026 History
1,376 lines · 66.6 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.6
4
5- Added search functionality to `/config` command for quickly filtering settings
6- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
7- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
8- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
9- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
10- Added an error display when the editor fails during Ctrl+G
11- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
12- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
13- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
14- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
15- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
16- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
17- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
18- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
19- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
20- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
21- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
22- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
23- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
24- Improved @ autocomplete with icons for different suggestion types and single-line formatting
25- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
26- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
27- Changed terminal title to "Claude Code" on startup for better window identification
28- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
29- [VSCode] Fixed usage indicator not updating after manual compact
30
31## 2.1.5
32
33- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
34
35## 2.1.4
36
37- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
38- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
39
40## 2.1.3
41
42- Merged slash commands and skills, simplifying the mental model with no change in behavior
43- Added release channel (`stable` or `latest`) toggle to `/config`
44- Added detection and warnings for unreachable permission rules, with warnings in `/doctor` and after saving rules that include the source of each rule and actionable fix guidance
45- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
46- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
47- Fixed mismatch between background task count in status bar and items shown in tasks dialog
48- Fixed sub-agents using the wrong model during conversation compaction
49- Fixed web search in sub-agents using incorrect model
50- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
51- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
52- Improved slash command suggestion readability by truncating long descriptions to 2 lines
53- Changed tool hook execution timeout from 60 seconds to 10 minutes
54- [VSCode] Added clickable destination selector for permission requests, allowing you to choose where settings are saved (this project, all projects, shared with team, or session only)
55
56## 2.1.2
57
58- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
59- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
60- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
61- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
62- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
63- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
64- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
65- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
66- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
67- Fixed updates incorrectly claiming another installation is in progress
68- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
69- Fixed remote session URL and teleport being broken when using `/tasks` command
70- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
71- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
72- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
73- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
74- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
75- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
76- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
77- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
78- [SDK] Changed minimum zod peer dependency to ^4.0.0
79- [VSCode] Fixed usage display not updating after manual compact
80
81## 2.1.0
82
83- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
84- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
85- Added support for `agent` field in skills to specify agent type for execution
86- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
87- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
88- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
89- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
90- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
91- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
92- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
93- Fixed Esc key with queued prompts to only move them to input without canceling the running task
94- Reduced permission prompts for complex bash commands
95- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
96- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
97- Fixed plugin path resolution for file-based marketplace sources
98- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
99- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
100- Fixed Claude in Chrome support for WSL environments
101- Fixed Windows native installer silently failing when executable creation fails
102- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
103- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
104- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
105- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
106- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
107- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
108- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
109- Added hooks support for skill and slash command frontmatter
110- Added new Vim motions: `;` and `,` to repeat f/F/t/T motions, `y` operator for yank with `yy`/`Y`, `p`/`P` for paste, text objects (`iw`, `aw`, `iW`, `aW`, `i"`, `a"`, `i'`, `a'`, `i(`, `a(`, `i[`, `a[`, `i{`, `a{`), `>>` and `<<` for indent/dedent, and `J` to join lines
111- Added `/plan` command shortcut to enable plan mode directly from the prompt
112- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
113- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
114- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
115- Added support for `once: true` config for hooks
116- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
117- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
118- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
119- Added left/right arrow key navigation for cycling through tabs in dialogs
120- Added real-time thinking block display in Ctrl+O transcript mode
121- Added filepath to full output in background bash task details dialog
122- Added Skills as a separate category in the context visualization
123- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
124- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
125- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
126- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
127- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
128- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
129- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
130- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
131- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
132- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
133- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
134- Fixed commands with `$()` command substitution failing with parse errors
135- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
136- Fixed bash command prefix extraction to correctly identify subcommands after global options (e.g., `git -C /path log` now correctly matches `Bash(git log:*)` rules)
137- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
138- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
139- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
140- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
141- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
142- Fixed subagents sometimes not inheriting the parent's model by default
143- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
144- Fixed duplicate Bash commands appearing in permission request option labels
145- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
146- Fixed background task completion notifications to appear proactively with bullet point
147- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
148- Fixed cursor disappearing after dismissing permission dialogs
149- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
150- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
151- Fixed images being silently dropped when queueing messages while backgrounding a task
152- Fixed large pasted images failing with "Image was too large" error
153- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
154- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
155- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
156- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
157- Fixed spinner token counter to properly accumulate tokens from subagents during execution
158- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
159- Fixed race condition where LSP tool could return "no server available" during startup
160- Fixed feedback submission hanging indefinitely when network requests timeout
161- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
162- Fixed hook success message showing trailing colon when hook has no output
163- Multiple optimizations to improve startup performance
164- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
165- Improved performance when reading Jupyter notebooks with many cells
166- Improved reliability for piped input like `cat refactor.md | claude`
167- Improved reliability for AskQuestion tool
168- Improved sed in-place edit commands to render as file edits with diff preview
169- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
170- Improved compaction reliability
171- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
172- Improved skills to show progress while executing, displaying tool uses as they happen
173- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
174- Improved skill suggestions to prioritize recently and frequently used skills
175- Improved spinner feedback when waiting for the first response token
176- Improved token count display in spinner to include tokens from background agents
177- Improved incremental output for async agents to give the main thread more control and visibility
178- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
179- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
180- Improved macOS screenshot paste reliability with TIFF format support
181- Improved `/stats` output
182- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
183- Changed "Interrupted" message color from red to grey for a less alarming appearance
184- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
185- Removed underline styling from image reference links
186- [SDK] Changed minimum zod peer dependency to ^4.0.0
187- [VSCode] Added currently selected model name to the context menu
188- [VSCode] Added descriptive labels on auto-accept permission button (e.g., "Yes, allow npm for this project" instead of "Yes, and don't ask again")
189- [VSCode] Fixed paragraph breaks not rendering in markdown content
190- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
191- [Windows] Fixed issue with improper rendering
192
193## 2.0.76
194
195- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
196
197## 2.0.75
198
199- Minor bugfixes
200
201## 2.0.74
202
203- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
204- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
205- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
206- Added syntax highlighting info to theme picker
207- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
208- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
209- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
210- Fixed a potential crash when syntax highlighting isn't initialized correctly
211- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
212- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
213- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
214- [Windows] Fixed issue with improper rendering
215- [VSCode] Added gift tag pictogram for year-end promotion message
216
217## 2.0.73
218
219- Added clickable `[Image #N]` links that open attached images in the default viewer
220- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
221- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
222- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
223- Fixed slow input history cycling and race condition that could overwrite text after message submission
224- Improved `/theme` command to open theme picker directly
225- Improved theme picker UI
226- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
227- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
228
229## 2.0.72
230
231- 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
232- Reduced terminal flickering
233- Added scannable QR code to mobile app tip for quick app downloads
234- Added loading indicator when resuming conversations for better feedback
235- Fixed `/context` command not respecting custom system prompts in non-interactive mode
236- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
237- Improved @ mention file suggestion speed (~3x faster in git repositories)
238- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
239- Improved settings validation errors to be more prominent
240- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
241
242## 2.0.71
243
244- Added /config toggle to enable/disable prompt suggestions
245- Added `/settings` as an alias for the `/config` command
246- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
247- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
248- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
249- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
250- New syntax highlighting engine for native build
251
252## 2.0.70
253
254- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
255- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
256- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
257- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
258- Fixed input being cleared when processing queued commands while the user was typing
259- Fixed prompt suggestions replacing typed input when pressing Tab
260- Fixed diff view not updating when terminal is resized
261- Improved memory usage by 3x for large conversations
262- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
263- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
264- Fix thinking mode toggle in /config not persisting correctly
265- Improve UI for file creation permission dialog
266
267## 2.0.69
268
269- Minor bugfixes
270
271## 2.0.68
272
273- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
274- Fixed a bug where disallowed MCP tools were visible to the model
275- Fixed an issue where steering messages could be lost while a subagent is working
276- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
277- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
278- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
279
280## 2.0.67
281
282- Thinking mode is now enabled by default for Opus 4.5
283- Thinking mode configuration has moved to /config
284- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
285- Show reason why autoupdater is disabled in `/doctor`
286- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
287- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
288- Fixed scroll position resetting after deleting a permission rule in `/permissions`
289- 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
290- Fixed `claude install --force` not bypassing stale lock files
291- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
292- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
293
294## 2.0.65
295
296- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
297- Added context window information to status line input
298- Added `fileSuggestion` setting for custom `@` file search commands
299- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
300- Fixed prompt not being saved to history when aborting a query with Escape
301- Fixed Read tool image handling to identify format from bytes instead of file extension
302
303## 2.0.64
304
305- Made auto-compacting instant
306- Agents and bash commands can run asynchronously and send messages to wake up the main agent
307- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
308- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
309- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
310- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
311- Fixed auto-loading .env when using native installer
312- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
313- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
314- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
315- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
316- Bedrock: Improve efficiency of token counting
317- Bedrock: Add support for `aws login` AWS Management Console credentials
318- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
319
320## 2.0.62
321
322- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
323- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
324- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
325- Fixed slash command selection not working when multiple commands share the same name
326- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
327- Fixed running versions getting removed because lock file incorrectly going stale
328- Fixed IDE diff tab not closing when rejecting file changes
329
330## 2.0.61
331
332- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
333
334## 2.0.60
335
336- Added background agent support. Agents run in the background while you work
337- Added --disable-slash-commands CLI flag to disable all slash commands
338- Added model name to "Co-Authored-By" commit messages
339- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
340- Updated Fetch to skip summarization for pre-approved websites
341- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
342
343## 2.0.59
344
345- Added --agent CLI flag to override the agent setting for the current session
346- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
347- VS Code: Fixed .claude.json config file being read from incorrect location
348
349## 2.0.58
350
351- Pro users now have access to Opus 4.5 as part of their subscription!
352- Fixed timer duration showing "11m 60s" instead of "12m 0s"
353- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
354
355## 2.0.57
356
357- Added feedback input when rejecting plans, allowing users to tell Claude what to change
358- VSCode: Added streaming message support for real-time response display
359
360## 2.0.56
361
362- Added setting to enable/disable terminal progress bar (OSC 9;4)
363- 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.
364
365## 2.0.55
366
367- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
368- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
369- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
370- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
371
372## 2.0.54
373
374- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
375- Fix issue with excessive iTerm notifications
376
377## 2.0.52
378
379- Fixed duplicate message display when starting Claude with a command line argument
380- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
381- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
382- Permit some uses of `$!` in bash commands
383
384## 2.0.51
385
386- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
387- Introducing Claude Code for Desktop: https://claude.com/download
388- 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
389- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
390- Plan Mode now builds more precise plans and executes more thoroughly
391- Usage limit notifications now easier to understand
392- Switched `/usage` back to "% used"
393- Fixed handling of thinking errors
394- Fixed performance regression
395
396## 2.0.50
397
398- Fixed bug preventing calling MCP tools that have nested references in their input schemas
399- Silenced a noisy but harmless error during upgrades
400- Improved ultrathink text display
401- Improved clarity of 5-hour session limit warning message
402
403## 2.0.49
404
405- Added readline-style ctrl-y for pasting deleted text
406- Improved clarity of usage limit warning message
407- Fixed handling of subagent permissions
408
409## 2.0.47
410
411- Improved error messages and validation for `claude --teleport`
412- Improved error handling in `/usage`
413- Fixed race condition with history entry not getting logged at exit
414- Fixed Vertex AI configuration not being applied from `settings.json`
415
416## 2.0.46
417
418- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
419
420## 2.0.45
421
422- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
423- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
424- Send background tasks to Claude Code on the web by starting a message with `&`
425
426## 2.0.43
427
428- Added `permissionMode` field for custom agents
429- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
430- Added skills frontmatter field to declare skills to auto-load for subagents
431- Added the `SubagentStart` hook event
432- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
433- Fixed duplicate rendering of some messages in the UI
434- Fixed some visual flickers
435- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
436
437## 2.0.42
438
439- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
440
441## 2.0.41
442
443- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
444- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
445- Fixed incorrect labeling of user settings vs project settings in command descriptions
446- Fixed crash when plugin command hooks timeout during execution
447- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
448- Fixed broken security documentation links in trust dialogs and onboarding
449- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
450- ctrl-r history search landing on a slash command no longer cancels the search
451- SDK: Support custom timeouts for hooks
452- Allow more safe git commands to run without approval
453- Plugins: Added support for sharing and installing output styles
454- Teleporting a session from web will automatically set the upstream branch
455
456## 2.0.37
457
458- Fixed how idleness is computed for notifications
459- Hooks: Added matcher values for Notification hook events
460- Output Styles: Added `keep-coding-instructions` option to frontmatter
461
462## 2.0.36
463
464- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
465- Fixed queued messages being incorrectly executed as bash commands
466- Fixed input being lost when typing while a queued message is processed
467
468## 2.0.35
469
470- Improve fuzzy search results when searching commands
471- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
472- 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
473- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
474- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
475
476## 2.0.34
477
478- VSCode Extension: Added setting to configure the initial permission mode for new conversations
479- Improved file path suggestion performance with native Rust-based fuzzy finder
480- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
481- Fixed memory crash when reading or writing large files (especially base64-encoded images)
482
483## 2.0.33
484
485- Native binary installs now launch quicker.
486- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
487- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
488
489## 2.0.32
490
491- Un-deprecate output styles based on community feedback
492- Added `companyAnnouncements` setting for displaying announcements on startup
493- Fixed hook progress messages not updating correctly during PostToolUse hook execution
494
495## 2.0.31
496
497- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
498- Vertex: add support for Web Search on supported models
499- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
500- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
501- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
502- Fixed plugin uninstall not removing plugins
503
504## 2.0.30
505
506- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
507- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
508- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
509- Added prompt-based stop hooks
510- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
511- Enabled SSE MCP servers on native build
512- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
513- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
514- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
515- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
516- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
517- Fixed bug that saved session permissions to local settings
518- Fixed MCP tools not being available to sub-agents
519- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
520- Fixed delay when navigating through typeahead suggestions with arrow keys
521- VSCode: Restored selection indicator in input footer showing current file or code selection status
522
523## 2.0.28
524
525- Plan mode: introduced new Plan subagent
526- Subagents: claude can now choose to resume subagents
527- Subagents: claude can dynamically choose the model used by its subagents
528- SDK: added --max-budget-usd flag
529- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
530- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
531- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
532- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
533- Various other bug fixes
534
535## 2.0.27
536
537- New UI for permission prompts
538- Added current branch filtering and search to session resume screen for easier navigation
539- Fixed directory @-mention causing "No assistant message found" error
540- VSCode Extension: Add config setting to include .gitignored files in file searches
541- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
542
543## 2.0.25
544
545- 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
546
547## 2.0.24
548
549- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
550- Claude Code Web: Support for Web -> CLI teleport
551- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
552- Bedrock: Display awsAuthRefresh output when auth is required
553
554## 2.0.22
555
556- Fixed content layout shift when scrolling through slash commands
557- IDE: Add toggle to enable/disable thinking.
558- Fix bug causing duplicate permission prompts with parallel tool calls
559- Add support for enterprise managed MCP allowlist and denylist
560
561## 2.0.21
562
563- Support MCP `structuredContent` field in tool responses
564- Added an interactive question tool
565- Claude will now ask you questions more often in plan mode
566- Added Haiku 4.5 as a model option for Pro users
567- Fixed an issue where queued commands don't have access to previous messages' output
568
569## 2.0.20
570
571- Added support for Claude Skills
572
573## 2.0.19
574
575- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
576- Fixed a bug where Haiku was unnecessarily called in print mode
577
578## 2.0.17
579
580- Added Haiku 4.5 to model selector!
581- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
582- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
583- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
584- OTEL: support HTTP_PROXY and HTTPS_PROXY
585- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
586
587## 2.0.15
588
589- Fixed bug with resuming where previously created files needed to be read again before writing
590- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
591
592## 2.0.14
593
594- Fix @-mentioning MCP servers to toggle them on/off
595- Improve permission checks for bash with inline env vars
596- Fix ultrathink + thinking toggle
597- Reduce unnecessary logins
598- Document --system-prompt
599- Several improvements to rendering
600- Plugins UI polish
601
602## 2.0.13
603
604- Fixed `/plugin` not working on native build
605
606## 2.0.12
607
608- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
609- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
610- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
611- `/plugin validate` command for validating plugin structure and configuration
612- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
613- Plugin documentation available at https://code.claude.com/docs/en/plugins
614- Comprehensive error messages and diagnostics via `/doctor` command
615- Avoid flickering in `/model` selector
616- Improvements to `/help`
617- Avoid mentioning hooks in `/resume` summaries
618- Changes to the "verbose" setting in `/config` now persist across sessions
619
620## 2.0.11
621
622- Reduced system prompt size by 1.4k tokens
623- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
624- Fixed Opus fallback rate limit errors appearing incorrectly
625- Fixed /add-dir command selecting wrong default tab
626
627## 2.0.10
628
629- Rewrote terminal renderer for buttery smooth UI
630- Enable/disable MCP servers by @mentioning, or in /mcp
631- Added tab completion for shell commands in bash mode
632- PreToolUse hooks can now modify tool inputs
633- Press Ctrl-G to edit your prompt in your system's configured text editor
634- Fixes for bash permission checks with environment variables in the command
635
636## 2.0.9
637
638- Fix regression where bash backgrounding stopped working
639
640## 2.0.8
641
642- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
643- IDE: Add drag-and-drop support for files and folders in chat
644- /context: Fix counting for thinking blocks
645- Improve message rendering for users with light themes on dark terminals
646- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
647
648## 2.0.5
649
650- IDE: Fix IME unintended message submission with Enter and Tab
651- IDE: Add "Open in Terminal" link in login screen
652- Fix unhandled OAuth expiration 401 API errors
653- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
654
655## 2.0.1
656
657- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
658- Various bug fixes and presentation improvements
659
660## 2.0.0
661
662- New native VS Code extension
663- Fresh coat of paint throughout the whole app
664- /rewind a conversation to undo code changes
665- /usage command to see plan limits
666- Tab to toggle thinking (sticky across sessions)
667- Ctrl-R to search history
668- Unshipped claude config command
669- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
670- SDK: The Claude Code SDK is now the Claude Agent SDK
671- Add subagents dynamically with `--agents` flag
672
673## 1.0.126
674
675- Enable /context command for Bedrock and Vertex
676- Add mTLS support for HTTP-based OpenTelemetry exporters
677
678## 1.0.124
679
680- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
681- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
682- No longer inform Claude of the list of allowed tools when permission is denied
683- Fixed security vulnerability in Bash tool permission checks
684- Improved VSCode extension performance for large files
685
686## 1.0.123
687
688- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
689- Fixed thinking mode triggering on negation phrases like "don't think"
690- Fixed rendering performance degradation during token streaming
691- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
692- Enhanced BashTool environment snapshot logging
693- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
694- Migrated --debug logging to a file, to enable easy tailing & filtering
695
696## 1.0.120
697
698- Fix input lag during typing, especially noticeable with large prompts
699- Improved VSCode extension command registry and sessions dialog user experience
700- Enhanced sessions dialog responsiveness and visual feedback
701- Fixed IDE compatibility issue by removing worktree support check
702- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
703
704## 1.0.119
705
706- Fix Windows issue where process visually freezes on entering interactive mode
707- Support dynamic headers for MCP servers via headersHelper configuration
708- Fix thinking mode not working in headless sessions
709- Fix slash commands now properly update allowed tools instead of replacing them
710
711## 1.0.117
712
713- Add Ctrl-R history search to recall previous commands like bash/zsh
714- Fix input lag while typing, especially on Windows
715- Add sed command to auto-allowed commands in acceptEdits mode
716- Fix Windows PATH comparison to be case-insensitive for drive letters
717- Add permissions management hint to /add-dir output
718
719## 1.0.115
720
721- Improve thinking mode display with enhanced visual effects
722- Type /t to temporarily disable thinking mode in your prompt
723- Improve path validation for glob and grep tools
724- Show condensed output for post-tool hooks to reduce visual clutter
725- Fix visual feedback when loading state completes
726- Improve UI consistency for permission request dialogs
727
728## 1.0.113
729
730- Deprecated piped input in interactive mode
731- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
732
733## 1.0.112
734
735- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
736- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
737- Hooks: Added systemMessage support for SessionEnd hooks
738- Added `spinnerTipsEnabled` setting to disable spinner tips
739- IDE: Various improvements and bug fixes
740
741## 1.0.111
742
743- /model now validates provided model names
744- Fixed Bash tool crashes caused by malformed shell syntax parsing
745
746## 1.0.110
747
748- /terminal-setup command now supports WezTerm
749- MCP: OAuth tokens now proactively refresh before expiration
750- Fixed reliability issues with background Bash processes
751
752## 1.0.109
753
754- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
755
756## 1.0.106
757
758- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
759
760## 1.0.97
761
762- Settings: /doctor now validates permission rule syntax and suggests corrections
763
764## 1.0.94
765
766- Vertex: add support for global endpoints for supported models
767- /memory command now allows direct editing of all imported memory files
768- SDK: Add custom tools as callbacks
769- Added /todos command to list current todo items
770
771## 1.0.93
772
773- Windows: Add alt + v shortcut for pasting images from clipboard
774- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
775
776## 1.0.90
777
778- Settings file changes take effect immediately - no restart required
779
780## 1.0.88
781
782- Fixed issue causing "OAuth authentication is currently not supported"
783- Status line input now includes `exceeds_200k_tokens`
784- Fixed incorrect usage tracking in /cost.
785- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
786- Bedrock: Updated default Sonnet model to Sonnet 4
787
788## 1.0.86
789
790- Added /context to help users self-serve debug context issues
791- SDK: Added UUID support for all SDK messages
792- SDK: Added `--replay-user-messages` to replay user messages back to stdout
793
794## 1.0.85
795
796- Status line input now includes session cost info
797- Hooks: Introduced SessionEnd hook
798
799## 1.0.84
800
801- Fix tool_use/tool_result id mismatch error when network is unstable
802- Fix Claude sometimes ignoring real-time steering when wrapping up a task
803- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
804- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
805
806## 1.0.83
807
808- @-mention: Support files with spaces in path
809- New shimmering spinner
810
811## 1.0.82
812
813- SDK: Add request cancellation support
814- SDK: New additionalDirectories option to search custom paths, improved slash command processing
815- Settings: Validation prevents invalid fields in .claude/settings.json files
816- MCP: Improve tool name consistency
817- Bash: Fix crash when Claude tries to automatically read large files
818
819## 1.0.81
820
821- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
822- Agents: Fix custom agent loading when agent files are unparsable
823
824## 1.0.80
825
826- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
827
828## 1.0.77
829
830- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
831- SDK: Add session support and permission denial tracking
832- Fix token limit errors in conversation summarization
833- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
834
835## 1.0.73
836
837- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
838- MCP: Press Esc to cancel OAuth authentication flows
839- Bash: Improved command validation and reduced false security warnings
840- UI: Enhanced spinner animations and status line visual hierarchy
841- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
842
843## 1.0.72
844
845- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
846
847## 1.0.71
848
849- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
850- Customizable status line: add your terminal prompt to Claude Code with /statusline
851
852## 1.0.70
853
854- Performance: Optimized message rendering for better performance with large contexts
855- Windows: Fixed native file search, ripgrep, and subagent functionality
856- Added support for @-mentions in slash command arguments
857
858## 1.0.69
859
860- Upgraded Opus to version 4.1
861
862## 1.0.68
863
864- Fix incorrect model names being used for certain commands like `/pr-comments`
865- 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.
866- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
867- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
868- SDK: Added canUseTool callback support for tool confirmation
869- Added `disableAllHooks` setting
870- Improved file suggestions performance in large repos
871
872## 1.0.65
873
874- IDE: Fixed connection stability issues and error handling for diagnostics
875- Windows: Fixed shell environment setup for users without .bashrc files
876
877## 1.0.64
878
879- Agents: Added model customization support - you can now specify which model an agent should use
880- Agents: Fixed unintended access to the recursive agent tool
881- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
882- SDK: Fixed user input tracking across multi-turn conversations
883- Added hidden files to file search and @-mention suggestions
884
885## 1.0.63
886
887- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
888
889## 1.0.62
890
891- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
892- Hooks: Added SessionStart hook for new session initialization
893- /add-dir command now supports typeahead for directory paths
894- Improved network connectivity check reliability
895
896## 1.0.61
897
898- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
899- Settings: Added `--settings` flag to load settings from a JSON file
900- Settings: Fixed resolution of settings files paths that are symlinks
901- OTEL: Fixed reporting of wrong organization after authentication changes
902- Slash commands: Fixed permissions checking for allowed-tools with Bash
903- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
904- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
905- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
906
907## 1.0.60
908
909- You can now create custom subagents for specialized tasks! Run /agents to get started
910
911## 1.0.59
912
913- SDK: Added tool confirmation support with canUseTool callback
914- SDK: Allow specifying env for spawned process
915- Hooks: Exposed PermissionDecision to hooks (including "ask")
916- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
917- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
918
919## 1.0.58
920
921- Added support for reading PDFs
922- MCP: Improved server health status display in 'claude mcp list'
923- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
924
925## 1.0.57
926
927- Added support for specifying a model in slash commands
928- Improved permission messages to help Claude understand allowed tools
929- Fix: Remove trailing newlines from bash output in terminal wrapping
930
931## 1.0.56
932
933- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
934- Fixes for WSL IDE detection
935- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
936
937## 1.0.55
938
939- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
940- Windows: fixed Ctrl+Z crash
941- SDK: Added ability to capture error logging
942- Add --system-prompt-file option to override system prompt in print mode
943
944## 1.0.54
945
946- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
947- Custom slash commands: Added argument-hint to frontmatter
948- Windows: OAuth uses port 45454 and properly constructs browser URL
949- Windows: mode switching now uses alt + m, and plan mode renders properly
950- Shell: Switch to in-memory shell snapshot to fix file-related errors
951
952## 1.0.53
953
954- Updated @-mention file truncation from 100 lines to 2000 lines
955- 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).
956
957## 1.0.52
958
959- Added support for MCP server instructions
960
961## 1.0.51
962
963- Added support for native Windows (requires Git for Windows)
964- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
965- Settings: /doctor can now help you identify and fix invalid setting files
966- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
967- Increased auto-compact warning threshold from 60% to 80%
968- Fixed an issue with handling user directories with spaces for shell snapshots
969- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
970- Custom slash commands: Fixed user-level commands in subdirectories
971- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
972
973## 1.0.48
974
975- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
976- Added progress messages to Bash tool based on the last 5 lines of command output
977- Added expanding variables support for MCP server configuration
978- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
979- Improved IDE extension path handling when Claude Code runs in WSL
980- Hooks: Added a PreCompact hook
981- Vim mode: Added c, f/F, t/T
982
983## 1.0.45
984
985- Redesigned Search (Grep) tool with new tool input parameters and features
986- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
987- Fixed config file corruption issue by enforcing atomic writes
988- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
989- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
990- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
991
992## 1.0.44
993
994- New /export command lets you quickly export a conversation for sharing
995- MCP: resource_link tool results are now supported
996- MCP: tool annotations and tool titles now display in /mcp view
997- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
998
999## 1.0.43
1000
1001- Fixed a bug where the theme selector was saving excessively
1002- Hooks: Added EPIPE system error handling
1003
1004## 1.0.42
1005
1006- Added tilde (`~`) expansion support to `/add-dir` command
1007
1008## 1.0.41
1009
1010- Hooks: Split Stop hook triggering into Stop and SubagentStop
1011- Hooks: Enabled optional timeout configuration for each command
1012- Hooks: Added "hook_event_name" to hook input
1013- Fixed a bug where MCP tools would display twice in tool list
1014- New tool parameters JSON for Bash tool in `tool_decision` event
1015
1016## 1.0.40
1017
1018- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1019
1020## 1.0.39
1021
1022- New Active Time metric in OpenTelemetry logging
1023
1024## 1.0.38
1025
1026- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1027
1028## 1.0.37
1029
1030- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1031
1032## 1.0.36
1033
1034- Web search now takes today's date into context
1035- Fixed a bug where stdio MCP servers were not terminating properly on exit
1036
1037## 1.0.35
1038
1039- Added support for MCP OAuth Authorization Server discovery
1040
1041## 1.0.34
1042
1043- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1044
1045## 1.0.33
1046
1047- Improved logging functionality with session ID support
1048- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1049- Improvements to plan mode
1050
1051## 1.0.32
1052
1053- Updated loopback config for litellm
1054- Added forceLoginMethod setting to bypass login selection screen
1055
1056## 1.0.31
1057
1058- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1059
1060## 1.0.30
1061
1062- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1063- Improved file path autocomplete with filename matching
1064- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1065- Enhanced jq regex support for complex filters with pipes and select
1066
1067## 1.0.29
1068
1069- Improved CJK character support in cursor navigation and rendering
1070
1071## 1.0.28
1072
1073- Slash commands: Fix selector display during history navigation
1074- Resizes images before upload to prevent API size limit errors
1075- Added XDG_CONFIG_HOME support to configuration directory
1076- Performance optimizations for memory usage
1077- New attributes (terminal.type, language) in OpenTelemetry logging
1078
1079## 1.0.27
1080
1081- Streamable HTTP MCP servers are now supported
1082- Remote MCP servers (SSE and HTTP) now support OAuth
1083- MCP resources can now be @-mentioned
1084- /resume slash command to switch conversations within Claude Code
1085
1086## 1.0.25
1087
1088- Slash commands: moved "project" and "user" prefixes to descriptions
1089- Slash commands: improved reliability for command discovery
1090- Improved support for Ghostty
1091- Improved web search reliability
1092
1093## 1.0.24
1094
1095- Improved /mcp output
1096- Fixed a bug where settings arrays got overwritten instead of merged
1097
1098## 1.0.23
1099
1100- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1101- Released Python SDK: pip install claude-code-sdk to get started
1102
1103## 1.0.22
1104
1105- SDK: Renamed `total_cost` to `total_cost_usd`
1106
1107## 1.0.21
1108
1109- Improved editing of files with tab-based indentation
1110- Fix for tool_use without matching tool_result errors
1111- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1112
1113## 1.0.18
1114
1115- Added --add-dir CLI argument for specifying additional working directories
1116- Added streaming input support without require -p flag
1117- Improved startup performance and session storage performance
1118- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1119- Added detailed MCP server tools display (/mcp)
1120- MCP authentication and permission improvements
1121- Added auto-reconnection for MCP SSE connections on disconnect
1122- Fixed issue where pasted content was lost when dialogs appeared
1123
1124## 1.0.17
1125
1126- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1127- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1128- MCP server list UI improvements
1129- Update Claude Code process title to display "claude" instead of "node"
1130
1131## 1.0.11
1132
1133- Claude Code can now also be used with a Claude Pro subscription
1134- Added /upgrade for smoother switching to Claude Max plans
1135- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1136- Improved shell configuration error handling
1137- Improved todo list handling during compaction
1138
1139## 1.0.10
1140
1141- Added markdown table support
1142- Improved streaming performance
1143
1144## 1.0.8
1145
1146- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1147- Increased default otel interval from 1s -> 5s
1148- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1149- Fixed a regression where search tools unnecessarily asked for permissions
1150- Added support for triggering thinking non-English languages
1151- Improved compacting UI
1152
1153## 1.0.7
1154
1155- Renamed /allowed-tools -> /permissions
1156- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1157- Deprecated claude config commands in favor of editing settings.json
1158- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1159- Improved error handling for /install-github-app
1160- Bugfixes, UI polish, and tool reliability improvements
1161
1162## 1.0.6
1163
1164- Improved edit reliability for tab-indented files
1165- Respect CLAUDE_CONFIG_DIR everywhere
1166- Reduced unnecessary tool permission prompts
1167- Added support for symlinks in @file typeahead
1168- Bugfixes, UI polish, and tool reliability improvements
1169
1170## 1.0.4
1171
1172- Fixed a bug where MCP tool errors weren't being parsed correctly
1173
1174## 1.0.1
1175
1176- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1177- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1178- Updated documentation links and OAuth process descriptions
1179
1180## 1.0.0
1181
1182- Claude Code is now generally available
1183- Introducing Sonnet 4 and Opus 4 models
1184
1185## 0.2.125
1186
1187- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1188- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1189
1190## 0.2.117
1191
1192- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1193- Introduced settings.cleanupPeriodDays
1194- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1195- Introduced --debug mode
1196
1197## 0.2.108
1198
1199- You can now send messages to Claude while it works to steer Claude in real-time
1200- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1201- Fixed a bug where thinking was not working in -p mode
1202- Fixed a regression in /cost reporting
1203- Deprecated MCP wizard interface in favor of other MCP commands
1204- Lots of other bugfixes and improvements
1205
1206## 0.2.107
1207
1208- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1209
1210## 0.2.106
1211
1212- MCP SSE server configs can now specify custom headers
1213- Fixed a bug where MCP permission prompt didn't always show correctly
1214
1215## 0.2.105
1216
1217- Claude can now search the web
1218- Moved system & account status to /status
1219- Added word movement keybindings for Vim
1220- Improved latency for startup, todo tool, and file edits
1221
1222## 0.2.102
1223
1224- Improved thinking triggering reliability
1225- Improved @mention reliability for images and folders
1226- You can now paste multiple large chunks into one prompt
1227
1228## 0.2.100
1229
1230- Fixed a crash caused by a stack overflow error
1231- Made db storage optional; missing db support disables --continue and --resume
1232
1233## 0.2.98
1234
1235- Fixed an issue where auto-compact was running twice
1236
1237## 0.2.96
1238
1239- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1240
1241## 0.2.93
1242
1243- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1244- Claude now has access to a Todo list that helps it stay on track and be more organized
1245
1246## 0.2.82
1247
1248- Added support for --disallowedTools
1249- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1250
1251## 0.2.75
1252
1253- Hit Enter to queue up additional messages while Claude is working
1254- Drag in or copy/paste image files directly into the prompt
1255- @-mention files to directly add them to context
1256- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1257- Improved performance for filename auto-complete
1258
1259## 0.2.74
1260
1261- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1262- Task tool can now perform writes and run bash commands
1263
1264## 0.2.72
1265
1266- Updated spinner to indicate tokens loaded and tool usage
1267
1268## 0.2.70
1269
1270- Network commands like curl are now available for Claude to use
1271- Claude can now run multiple web queries in parallel
1272- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1273
1274## 0.2.69
1275
1276- Fixed UI glitches with improved Select component behavior
1277- Enhanced terminal output display with better text truncation logic
1278
1279## 0.2.67
1280
1281- Shared project permission rules can be saved in .claude/settings.json
1282
1283## 0.2.66
1284
1285- Print mode (-p) now supports streaming output via --output-format=stream-json
1286- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1287
1288## 0.2.63
1289
1290- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1291
1292## 0.2.61
1293
1294- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1295- Enhanced image detection for more reliable clipboard paste functionality
1296- Fixed an issue where ESC key could crash the conversation history selector
1297
1298## 0.2.59
1299
1300- Copy+paste images directly into your prompt
1301- Improved progress indicators for bash and fetch tools
1302- Bugfixes for non-interactive mode (-p)
1303
1304## 0.2.54
1305
1306- Quickly add to Memory by starting your message with '#'
1307- Press ctrl+r to see full output for long tool results
1308- Added support for MCP SSE transport
1309
1310## 0.2.53
1311
1312- New web fetch tool lets Claude view URLs that you paste in
1313- Fixed a bug with JPEG detection
1314
1315## 0.2.50
1316
1317- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1318
1319## 0.2.49
1320
1321- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1322
1323## 0.2.47
1324
1325- Press Tab to auto-complete file and folder names
1326- Press Shift + Tab to toggle auto-accept for file edits
1327- Automatic conversation compaction for infinite conversation length (toggle with /config)
1328
1329## 0.2.44
1330
1331- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1332
1333## 0.2.41
1334
1335- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1336- MCP server startup no longer blocks the app from starting up
1337
1338## 0.2.37
1339
1340- New /release-notes command lets you view release notes at any time
1341- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1342
1343## 0.2.36
1344
1345- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1346- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1347
1348## 0.2.34
1349
1350- Vim bindings for text input - enable with /vim or /config
1351
1352## 0.2.32
1353
1354- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1355- Fix for some PersistentShell issues
1356
1357## 0.2.31
1358
1359- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1360- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1361
1362## 0.2.30
1363
1364- Added ANSI color theme for better terminal compatibility
1365- Fixed issue where slash command arguments weren't being sent properly
1366- (Mac-only) API keys are now stored in macOS Keychain
1367
1368## 0.2.26
1369
1370- New /approved-tools command for managing tool permissions
1371- Word-level diff display for improved code readability
1372- Fuzzy matching for slash commands
1373
1374## 0.2.21
1375
1376- Fuzzy matching for /commands