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 0359f24 · on Jan 11, 2026 History
1,344 lines · 64.0 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.4
4
5- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
6- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
7
8## 2.1.3
9
10- Merged slash commands and skills, simplifying the mental model with no change in behavior
11- Added release channel (`stable` or `latest`) toggle to `/config`
12- 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
13- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
14- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
15- Fixed mismatch between background task count in status bar and items shown in tasks dialog
16- Fixed sub-agents using the wrong model during conversation compaction
17- Fixed web search in sub-agents using incorrect model
18- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
19- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
20- Improved slash command suggestion readability by truncating long descriptions to 2 lines
21- Changed tool hook execution timeout from 60 seconds to 10 minutes
22- [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)
23
24## 2.1.2
25
26- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
27- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
28- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
29- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
30- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
31- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
32- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
33- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
34- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
35- Fixed updates incorrectly claiming another installation is in progress
36- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
37- Fixed remote session URL and teleport being broken when using `/tasks` command
38- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
39- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
40- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
41- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
42- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
43- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
44- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
45- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
46- [SDK] Changed minimum zod peer dependency to ^4.0.0
47- [VSCode] Fixed usage display not updating after manual compact
48
49## 2.1.0
50
51- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
52- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
53- Added support for `agent` field in skills to specify agent type for execution
54- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
55- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
56- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
57- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
58- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
59- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
60- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
61- Fixed Esc key with queued prompts to only move them to input without canceling the running task
62- Reduced permission prompts for complex bash commands
63- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
64- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
65- Fixed plugin path resolution for file-based marketplace sources
66- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
67- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
68- Fixed Claude in Chrome support for WSL environments
69- Fixed Windows native installer silently failing when executable creation fails
70- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
71- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
72- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
73- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
74- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
75- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
76- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
77- Added hooks support for skill and slash command frontmatter
78- 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
79- Added `/plan` command shortcut to enable plan mode directly from the prompt
80- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
81- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
82- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
83- Added support for `once: true` config for hooks
84- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
85- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
86- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
87- Added left/right arrow key navigation for cycling through tabs in dialogs
88- Added real-time thinking block display in Ctrl+O transcript mode
89- Added filepath to full output in background bash task details dialog
90- Added Skills as a separate category in the context visualization
91- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
92- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
93- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
94- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
95- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
96- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
97- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
98- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
99- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
100- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
101- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
102- Fixed commands with `$()` command substitution failing with parse errors
103- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
104- 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)
105- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
106- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
107- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
108- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
109- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
110- Fixed subagents sometimes not inheriting the parent's model by default
111- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
112- Fixed duplicate Bash commands appearing in permission request option labels
113- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
114- Fixed background task completion notifications to appear proactively with bullet point
115- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
116- Fixed cursor disappearing after dismissing permission dialogs
117- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
118- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
119- Fixed images being silently dropped when queueing messages while backgrounding a task
120- Fixed large pasted images failing with "Image was too large" error
121- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
122- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
123- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
124- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
125- Fixed spinner token counter to properly accumulate tokens from subagents during execution
126- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
127- Fixed race condition where LSP tool could return "no server available" during startup
128- Fixed feedback submission hanging indefinitely when network requests timeout
129- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
130- Fixed hook success message showing trailing colon when hook has no output
131- Multiple optimizations to improve startup performance
132- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
133- Improved performance when reading Jupyter notebooks with many cells
134- Improved reliability for piped input like `cat refactor.md | claude`
135- Improved reliability for AskQuestion tool
136- Improved sed in-place edit commands to render as file edits with diff preview
137- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
138- Improved compaction reliability
139- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
140- Improved skills to show progress while executing, displaying tool uses as they happen
141- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
142- Improved skill suggestions to prioritize recently and frequently used skills
143- Improved spinner feedback when waiting for the first response token
144- Improved token count display in spinner to include tokens from background agents
145- Improved incremental output for async agents to give the main thread more control and visibility
146- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
147- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
148- Improved macOS screenshot paste reliability with TIFF format support
149- Improved `/stats` output
150- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
151- Changed "Interrupted" message color from red to grey for a less alarming appearance
152- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
153- Removed underline styling from image reference links
154- [SDK] Changed minimum zod peer dependency to ^4.0.0
155- [VSCode] Added currently selected model name to the context menu
156- [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")
157- [VSCode] Fixed paragraph breaks not rendering in markdown content
158- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
159- [Windows] Fixed issue with improper rendering
160
161## 2.0.76
162
163- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
164
165## 2.0.75
166
167- Minor bugfixes
168
169## 2.0.74
170
171- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
172- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
173- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
174- Added syntax highlighting info to theme picker
175- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
176- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
177- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
178- Fixed a potential crash when syntax highlighting isn't initialized correctly
179- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
180- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
181- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
182- [Windows] Fixed issue with improper rendering
183- [VSCode] Added gift tag pictogram for year-end promotion message
184
185## 2.0.73
186
187- Added clickable `[Image #N]` links that open attached images in the default viewer
188- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
189- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
190- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
191- Fixed slow input history cycling and race condition that could overwrite text after message submission
192- Improved `/theme` command to open theme picker directly
193- Improved theme picker UI
194- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
195- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
196
197## 2.0.72
198
199- 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
200- Reduced terminal flickering
201- Added scannable QR code to mobile app tip for quick app downloads
202- Added loading indicator when resuming conversations for better feedback
203- Fixed `/context` command not respecting custom system prompts in non-interactive mode
204- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
205- Improved @ mention file suggestion speed (~3x faster in git repositories)
206- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
207- Improved settings validation errors to be more prominent
208- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
209
210## 2.0.71
211
212- Added /config toggle to enable/disable prompt suggestions
213- Added `/settings` as an alias for the `/config` command
214- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
215- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
216- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
217- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
218- New syntax highlighting engine for native build
219
220## 2.0.70
221
222- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
223- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
224- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
225- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
226- Fixed input being cleared when processing queued commands while the user was typing
227- Fixed prompt suggestions replacing typed input when pressing Tab
228- Fixed diff view not updating when terminal is resized
229- Improved memory usage by 3x for large conversations
230- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
231- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
232- Fix thinking mode toggle in /config not persisting correctly
233- Improve UI for file creation permission dialog
234
235## 2.0.69
236
237- Minor bugfixes
238
239## 2.0.68
240
241- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
242- Fixed a bug where disallowed MCP tools were visible to the model
243- Fixed an issue where steering messages could be lost while a subagent is working
244- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
245- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
246- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
247
248## 2.0.67
249
250- Thinking mode is now enabled by default for Opus 4.5
251- Thinking mode configuration has moved to /config
252- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
253- Show reason why autoupdater is disabled in `/doctor`
254- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
255- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
256- Fixed scroll position resetting after deleting a permission rule in `/permissions`
257- 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
258- Fixed `claude install --force` not bypassing stale lock files
259- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
260- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
261
262## 2.0.65
263
264- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
265- Added context window information to status line input
266- Added `fileSuggestion` setting for custom `@` file search commands
267- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
268- Fixed prompt not being saved to history when aborting a query with Escape
269- Fixed Read tool image handling to identify format from bytes instead of file extension
270
271## 2.0.64
272
273- Made auto-compacting instant
274- Agents and bash commands can run asynchronously and send messages to wake up the main agent
275- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
276- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
277- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
278- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
279- Fixed auto-loading .env when using native installer
280- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
281- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
282- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
283- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
284- Bedrock: Improve efficiency of token counting
285- Bedrock: Add support for `aws login` AWS Management Console credentials
286- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
287
288## 2.0.62
289
290- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
291- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
292- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
293- Fixed slash command selection not working when multiple commands share the same name
294- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
295- Fixed running versions getting removed because lock file incorrectly going stale
296- Fixed IDE diff tab not closing when rejecting file changes
297
298## 2.0.61
299
300- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
301
302## 2.0.60
303
304- Added background agent support. Agents run in the background while you work
305- Added --disable-slash-commands CLI flag to disable all slash commands
306- Added model name to "Co-Authored-By" commit messages
307- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
308- Updated Fetch to skip summarization for pre-approved websites
309- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
310
311## 2.0.59
312
313- Added --agent CLI flag to override the agent setting for the current session
314- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
315- VS Code: Fixed .claude.json config file being read from incorrect location
316
317## 2.0.58
318
319- Pro users now have access to Opus 4.5 as part of their subscription!
320- Fixed timer duration showing "11m 60s" instead of "12m 0s"
321- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
322
323## 2.0.57
324
325- Added feedback input when rejecting plans, allowing users to tell Claude what to change
326- VSCode: Added streaming message support for real-time response display
327
328## 2.0.56
329
330- Added setting to enable/disable terminal progress bar (OSC 9;4)
331- 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.
332
333## 2.0.55
334
335- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
336- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
337- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
338- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
339
340## 2.0.54
341
342- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
343- Fix issue with excessive iTerm notifications
344
345## 2.0.52
346
347- Fixed duplicate message display when starting Claude with a command line argument
348- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
349- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
350- Permit some uses of `$!` in bash commands
351
352## 2.0.51
353
354- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
355- Introducing Claude Code for Desktop: https://claude.com/download
356- 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
357- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
358- Plan Mode now builds more precise plans and executes more thoroughly
359- Usage limit notifications now easier to understand
360- Switched `/usage` back to "% used"
361- Fixed handling of thinking errors
362- Fixed performance regression
363
364## 2.0.50
365
366- Fixed bug preventing calling MCP tools that have nested references in their input schemas
367- Silenced a noisy but harmless error during upgrades
368- Improved ultrathink text display
369- Improved clarity of 5-hour session limit warning message
370
371## 2.0.49
372
373- Added readline-style ctrl-y for pasting deleted text
374- Improved clarity of usage limit warning message
375- Fixed handling of subagent permissions
376
377## 2.0.47
378
379- Improved error messages and validation for `claude --teleport`
380- Improved error handling in `/usage`
381- Fixed race condition with history entry not getting logged at exit
382- Fixed Vertex AI configuration not being applied from `settings.json`
383
384## 2.0.46
385
386- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
387
388## 2.0.45
389
390- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
391- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
392- Send background tasks to Claude Code on the web by starting a message with `&`
393
394## 2.0.43
395
396- Added `permissionMode` field for custom agents
397- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
398- Added skills frontmatter field to declare skills to auto-load for subagents
399- Added the `SubagentStart` hook event
400- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
401- Fixed duplicate rendering of some messages in the UI
402- Fixed some visual flickers
403- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
404
405## 2.0.42
406
407- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
408
409## 2.0.41
410
411- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
412- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
413- Fixed incorrect labeling of user settings vs project settings in command descriptions
414- Fixed crash when plugin command hooks timeout during execution
415- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
416- Fixed broken security documentation links in trust dialogs and onboarding
417- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
418- ctrl-r history search landing on a slash command no longer cancels the search
419- SDK: Support custom timeouts for hooks
420- Allow more safe git commands to run without approval
421- Plugins: Added support for sharing and installing output styles
422- Teleporting a session from web will automatically set the upstream branch
423
424## 2.0.37
425
426- Fixed how idleness is computed for notifications
427- Hooks: Added matcher values for Notification hook events
428- Output Styles: Added `keep-coding-instructions` option to frontmatter
429
430## 2.0.36
431
432- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
433- Fixed queued messages being incorrectly executed as bash commands
434- Fixed input being lost when typing while a queued message is processed
435
436## 2.0.35
437
438- Improve fuzzy search results when searching commands
439- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
440- 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
441- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
442- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
443
444## 2.0.34
445
446- VSCode Extension: Added setting to configure the initial permission mode for new conversations
447- Improved file path suggestion performance with native Rust-based fuzzy finder
448- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
449- Fixed memory crash when reading or writing large files (especially base64-encoded images)
450
451## 2.0.33
452
453- Native binary installs now launch quicker.
454- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
455- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
456
457## 2.0.32
458
459- Un-deprecate output styles based on community feedback
460- Added `companyAnnouncements` setting for displaying announcements on startup
461- Fixed hook progress messages not updating correctly during PostToolUse hook execution
462
463## 2.0.31
464
465- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
466- Vertex: add support for Web Search on supported models
467- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
468- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
469- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
470- Fixed plugin uninstall not removing plugins
471
472## 2.0.30
473
474- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
475- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
476- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
477- Added prompt-based stop hooks
478- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
479- Enabled SSE MCP servers on native build
480- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
481- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
482- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
483- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
484- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
485- Fixed bug that saved session permissions to local settings
486- Fixed MCP tools not being available to sub-agents
487- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
488- Fixed delay when navigating through typeahead suggestions with arrow keys
489- VSCode: Restored selection indicator in input footer showing current file or code selection status
490
491## 2.0.28
492
493- Plan mode: introduced new Plan subagent
494- Subagents: claude can now choose to resume subagents
495- Subagents: claude can dynamically choose the model used by its subagents
496- SDK: added --max-budget-usd flag
497- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
498- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
499- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
500- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
501- Various other bug fixes
502
503## 2.0.27
504
505- New UI for permission prompts
506- Added current branch filtering and search to session resume screen for easier navigation
507- Fixed directory @-mention causing "No assistant message found" error
508- VSCode Extension: Add config setting to include .gitignored files in file searches
509- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
510
511## 2.0.25
512
513- 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
514
515## 2.0.24
516
517- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
518- Claude Code Web: Support for Web -> CLI teleport
519- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
520- Bedrock: Display awsAuthRefresh output when auth is required
521
522## 2.0.22
523
524- Fixed content layout shift when scrolling through slash commands
525- IDE: Add toggle to enable/disable thinking.
526- Fix bug causing duplicate permission prompts with parallel tool calls
527- Add support for enterprise managed MCP allowlist and denylist
528
529## 2.0.21
530
531- Support MCP `structuredContent` field in tool responses
532- Added an interactive question tool
533- Claude will now ask you questions more often in plan mode
534- Added Haiku 4.5 as a model option for Pro users
535- Fixed an issue where queued commands don't have access to previous messages' output
536
537## 2.0.20
538
539- Added support for Claude Skills
540
541## 2.0.19
542
543- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
544- Fixed a bug where Haiku was unnecessarily called in print mode
545
546## 2.0.17
547
548- Added Haiku 4.5 to model selector!
549- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
550- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
551- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
552- OTEL: support HTTP_PROXY and HTTPS_PROXY
553- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
554
555## 2.0.15
556
557- Fixed bug with resuming where previously created files needed to be read again before writing
558- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
559
560## 2.0.14
561
562- Fix @-mentioning MCP servers to toggle them on/off
563- Improve permission checks for bash with inline env vars
564- Fix ultrathink + thinking toggle
565- Reduce unnecessary logins
566- Document --system-prompt
567- Several improvements to rendering
568- Plugins UI polish
569
570## 2.0.13
571
572- Fixed `/plugin` not working on native build
573
574## 2.0.12
575
576- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
577- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
578- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
579- `/plugin validate` command for validating plugin structure and configuration
580- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
581- Plugin documentation available at https://code.claude.com/docs/en/plugins
582- Comprehensive error messages and diagnostics via `/doctor` command
583- Avoid flickering in `/model` selector
584- Improvements to `/help`
585- Avoid mentioning hooks in `/resume` summaries
586- Changes to the "verbose" setting in `/config` now persist across sessions
587
588## 2.0.11
589
590- Reduced system prompt size by 1.4k tokens
591- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
592- Fixed Opus fallback rate limit errors appearing incorrectly
593- Fixed /add-dir command selecting wrong default tab
594
595## 2.0.10
596
597- Rewrote terminal renderer for buttery smooth UI
598- Enable/disable MCP servers by @mentioning, or in /mcp
599- Added tab completion for shell commands in bash mode
600- PreToolUse hooks can now modify tool inputs
601- Press Ctrl-G to edit your prompt in your system's configured text editor
602- Fixes for bash permission checks with environment variables in the command
603
604## 2.0.9
605
606- Fix regression where bash backgrounding stopped working
607
608## 2.0.8
609
610- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
611- IDE: Add drag-and-drop support for files and folders in chat
612- /context: Fix counting for thinking blocks
613- Improve message rendering for users with light themes on dark terminals
614- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
615
616## 2.0.5
617
618- IDE: Fix IME unintended message submission with Enter and Tab
619- IDE: Add "Open in Terminal" link in login screen
620- Fix unhandled OAuth expiration 401 API errors
621- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
622
623## 2.0.1
624
625- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
626- Various bug fixes and presentation improvements
627
628## 2.0.0
629
630- New native VS Code extension
631- Fresh coat of paint throughout the whole app
632- /rewind a conversation to undo code changes
633- /usage command to see plan limits
634- Tab to toggle thinking (sticky across sessions)
635- Ctrl-R to search history
636- Unshipped claude config command
637- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
638- SDK: The Claude Code SDK is now the Claude Agent SDK
639- Add subagents dynamically with `--agents` flag
640
641## 1.0.126
642
643- Enable /context command for Bedrock and Vertex
644- Add mTLS support for HTTP-based OpenTelemetry exporters
645
646## 1.0.124
647
648- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
649- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
650- No longer inform Claude of the list of allowed tools when permission is denied
651- Fixed security vulnerability in Bash tool permission checks
652- Improved VSCode extension performance for large files
653
654## 1.0.123
655
656- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
657- Fixed thinking mode triggering on negation phrases like "don't think"
658- Fixed rendering performance degradation during token streaming
659- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
660- Enhanced BashTool environment snapshot logging
661- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
662- Migrated --debug logging to a file, to enable easy tailing & filtering
663
664## 1.0.120
665
666- Fix input lag during typing, especially noticeable with large prompts
667- Improved VSCode extension command registry and sessions dialog user experience
668- Enhanced sessions dialog responsiveness and visual feedback
669- Fixed IDE compatibility issue by removing worktree support check
670- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
671
672## 1.0.119
673
674- Fix Windows issue where process visually freezes on entering interactive mode
675- Support dynamic headers for MCP servers via headersHelper configuration
676- Fix thinking mode not working in headless sessions
677- Fix slash commands now properly update allowed tools instead of replacing them
678
679## 1.0.117
680
681- Add Ctrl-R history search to recall previous commands like bash/zsh
682- Fix input lag while typing, especially on Windows
683- Add sed command to auto-allowed commands in acceptEdits mode
684- Fix Windows PATH comparison to be case-insensitive for drive letters
685- Add permissions management hint to /add-dir output
686
687## 1.0.115
688
689- Improve thinking mode display with enhanced visual effects
690- Type /t to temporarily disable thinking mode in your prompt
691- Improve path validation for glob and grep tools
692- Show condensed output for post-tool hooks to reduce visual clutter
693- Fix visual feedback when loading state completes
694- Improve UI consistency for permission request dialogs
695
696## 1.0.113
697
698- Deprecated piped input in interactive mode
699- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
700
701## 1.0.112
702
703- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
704- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
705- Hooks: Added systemMessage support for SessionEnd hooks
706- Added `spinnerTipsEnabled` setting to disable spinner tips
707- IDE: Various improvements and bug fixes
708
709## 1.0.111
710
711- /model now validates provided model names
712- Fixed Bash tool crashes caused by malformed shell syntax parsing
713
714## 1.0.110
715
716- /terminal-setup command now supports WezTerm
717- MCP: OAuth tokens now proactively refresh before expiration
718- Fixed reliability issues with background Bash processes
719
720## 1.0.109
721
722- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
723
724## 1.0.106
725
726- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
727
728## 1.0.97
729
730- Settings: /doctor now validates permission rule syntax and suggests corrections
731
732## 1.0.94
733
734- Vertex: add support for global endpoints for supported models
735- /memory command now allows direct editing of all imported memory files
736- SDK: Add custom tools as callbacks
737- Added /todos command to list current todo items
738
739## 1.0.93
740
741- Windows: Add alt + v shortcut for pasting images from clipboard
742- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
743
744## 1.0.90
745
746- Settings file changes take effect immediately - no restart required
747
748## 1.0.88
749
750- Fixed issue causing "OAuth authentication is currently not supported"
751- Status line input now includes `exceeds_200k_tokens`
752- Fixed incorrect usage tracking in /cost.
753- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
754- Bedrock: Updated default Sonnet model to Sonnet 4
755
756## 1.0.86
757
758- Added /context to help users self-serve debug context issues
759- SDK: Added UUID support for all SDK messages
760- SDK: Added `--replay-user-messages` to replay user messages back to stdout
761
762## 1.0.85
763
764- Status line input now includes session cost info
765- Hooks: Introduced SessionEnd hook
766
767## 1.0.84
768
769- Fix tool_use/tool_result id mismatch error when network is unstable
770- Fix Claude sometimes ignoring real-time steering when wrapping up a task
771- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
772- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
773
774## 1.0.83
775
776- @-mention: Support files with spaces in path
777- New shimmering spinner
778
779## 1.0.82
780
781- SDK: Add request cancellation support
782- SDK: New additionalDirectories option to search custom paths, improved slash command processing
783- Settings: Validation prevents invalid fields in .claude/settings.json files
784- MCP: Improve tool name consistency
785- Bash: Fix crash when Claude tries to automatically read large files
786
787## 1.0.81
788
789- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
790- Agents: Fix custom agent loading when agent files are unparsable
791
792## 1.0.80
793
794- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
795
796## 1.0.77
797
798- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
799- SDK: Add session support and permission denial tracking
800- Fix token limit errors in conversation summarization
801- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
802
803## 1.0.73
804
805- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
806- MCP: Press Esc to cancel OAuth authentication flows
807- Bash: Improved command validation and reduced false security warnings
808- UI: Enhanced spinner animations and status line visual hierarchy
809- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
810
811## 1.0.72
812
813- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
814
815## 1.0.71
816
817- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
818- Customizable status line: add your terminal prompt to Claude Code with /statusline
819
820## 1.0.70
821
822- Performance: Optimized message rendering for better performance with large contexts
823- Windows: Fixed native file search, ripgrep, and subagent functionality
824- Added support for @-mentions in slash command arguments
825
826## 1.0.69
827
828- Upgraded Opus to version 4.1
829
830## 1.0.68
831
832- Fix incorrect model names being used for certain commands like `/pr-comments`
833- 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.
834- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
835- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
836- SDK: Added canUseTool callback support for tool confirmation
837- Added `disableAllHooks` setting
838- Improved file suggestions performance in large repos
839
840## 1.0.65
841
842- IDE: Fixed connection stability issues and error handling for diagnostics
843- Windows: Fixed shell environment setup for users without .bashrc files
844
845## 1.0.64
846
847- Agents: Added model customization support - you can now specify which model an agent should use
848- Agents: Fixed unintended access to the recursive agent tool
849- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
850- SDK: Fixed user input tracking across multi-turn conversations
851- Added hidden files to file search and @-mention suggestions
852
853## 1.0.63
854
855- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
856
857## 1.0.62
858
859- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
860- Hooks: Added SessionStart hook for new session initialization
861- /add-dir command now supports typeahead for directory paths
862- Improved network connectivity check reliability
863
864## 1.0.61
865
866- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
867- Settings: Added `--settings` flag to load settings from a JSON file
868- Settings: Fixed resolution of settings files paths that are symlinks
869- OTEL: Fixed reporting of wrong organization after authentication changes
870- Slash commands: Fixed permissions checking for allowed-tools with Bash
871- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
872- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
873- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
874
875## 1.0.60
876
877- You can now create custom subagents for specialized tasks! Run /agents to get started
878
879## 1.0.59
880
881- SDK: Added tool confirmation support with canUseTool callback
882- SDK: Allow specifying env for spawned process
883- Hooks: Exposed PermissionDecision to hooks (including "ask")
884- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
885- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
886
887## 1.0.58
888
889- Added support for reading PDFs
890- MCP: Improved server health status display in 'claude mcp list'
891- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
892
893## 1.0.57
894
895- Added support for specifying a model in slash commands
896- Improved permission messages to help Claude understand allowed tools
897- Fix: Remove trailing newlines from bash output in terminal wrapping
898
899## 1.0.56
900
901- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
902- Fixes for WSL IDE detection
903- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
904
905## 1.0.55
906
907- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
908- Windows: fixed Ctrl+Z crash
909- SDK: Added ability to capture error logging
910- Add --system-prompt-file option to override system prompt in print mode
911
912## 1.0.54
913
914- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
915- Custom slash commands: Added argument-hint to frontmatter
916- Windows: OAuth uses port 45454 and properly constructs browser URL
917- Windows: mode switching now uses alt + m, and plan mode renders properly
918- Shell: Switch to in-memory shell snapshot to fix file-related errors
919
920## 1.0.53
921
922- Updated @-mention file truncation from 100 lines to 2000 lines
923- 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).
924
925## 1.0.52
926
927- Added support for MCP server instructions
928
929## 1.0.51
930
931- Added support for native Windows (requires Git for Windows)
932- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
933- Settings: /doctor can now help you identify and fix invalid setting files
934- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
935- Increased auto-compact warning threshold from 60% to 80%
936- Fixed an issue with handling user directories with spaces for shell snapshots
937- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
938- Custom slash commands: Fixed user-level commands in subdirectories
939- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
940
941## 1.0.48
942
943- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
944- Added progress messages to Bash tool based on the last 5 lines of command output
945- Added expanding variables support for MCP server configuration
946- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
947- Improved IDE extension path handling when Claude Code runs in WSL
948- Hooks: Added a PreCompact hook
949- Vim mode: Added c, f/F, t/T
950
951## 1.0.45
952
953- Redesigned Search (Grep) tool with new tool input parameters and features
954- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
955- Fixed config file corruption issue by enforcing atomic writes
956- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
957- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
958- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
959
960## 1.0.44
961
962- New /export command lets you quickly export a conversation for sharing
963- MCP: resource_link tool results are now supported
964- MCP: tool annotations and tool titles now display in /mcp view
965- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
966
967## 1.0.43
968
969- Fixed a bug where the theme selector was saving excessively
970- Hooks: Added EPIPE system error handling
971
972## 1.0.42
973
974- Added tilde (`~`) expansion support to `/add-dir` command
975
976## 1.0.41
977
978- Hooks: Split Stop hook triggering into Stop and SubagentStop
979- Hooks: Enabled optional timeout configuration for each command
980- Hooks: Added "hook_event_name" to hook input
981- Fixed a bug where MCP tools would display twice in tool list
982- New tool parameters JSON for Bash tool in `tool_decision` event
983
984## 1.0.40
985
986- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
987
988## 1.0.39
989
990- New Active Time metric in OpenTelemetry logging
991
992## 1.0.38
993
994- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
995
996## 1.0.37
997
998- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
999
1000## 1.0.36
1001
1002- Web search now takes today's date into context
1003- Fixed a bug where stdio MCP servers were not terminating properly on exit
1004
1005## 1.0.35
1006
1007- Added support for MCP OAuth Authorization Server discovery
1008
1009## 1.0.34
1010
1011- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1012
1013## 1.0.33
1014
1015- Improved logging functionality with session ID support
1016- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1017- Improvements to plan mode
1018
1019## 1.0.32
1020
1021- Updated loopback config for litellm
1022- Added forceLoginMethod setting to bypass login selection screen
1023
1024## 1.0.31
1025
1026- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1027
1028## 1.0.30
1029
1030- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1031- Improved file path autocomplete with filename matching
1032- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1033- Enhanced jq regex support for complex filters with pipes and select
1034
1035## 1.0.29
1036
1037- Improved CJK character support in cursor navigation and rendering
1038
1039## 1.0.28
1040
1041- Slash commands: Fix selector display during history navigation
1042- Resizes images before upload to prevent API size limit errors
1043- Added XDG_CONFIG_HOME support to configuration directory
1044- Performance optimizations for memory usage
1045- New attributes (terminal.type, language) in OpenTelemetry logging
1046
1047## 1.0.27
1048
1049- Streamable HTTP MCP servers are now supported
1050- Remote MCP servers (SSE and HTTP) now support OAuth
1051- MCP resources can now be @-mentioned
1052- /resume slash command to switch conversations within Claude Code
1053
1054## 1.0.25
1055
1056- Slash commands: moved "project" and "user" prefixes to descriptions
1057- Slash commands: improved reliability for command discovery
1058- Improved support for Ghostty
1059- Improved web search reliability
1060
1061## 1.0.24
1062
1063- Improved /mcp output
1064- Fixed a bug where settings arrays got overwritten instead of merged
1065
1066## 1.0.23
1067
1068- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1069- Released Python SDK: pip install claude-code-sdk to get started
1070
1071## 1.0.22
1072
1073- SDK: Renamed `total_cost` to `total_cost_usd`
1074
1075## 1.0.21
1076
1077- Improved editing of files with tab-based indentation
1078- Fix for tool_use without matching tool_result errors
1079- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1080
1081## 1.0.18
1082
1083- Added --add-dir CLI argument for specifying additional working directories
1084- Added streaming input support without require -p flag
1085- Improved startup performance and session storage performance
1086- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1087- Added detailed MCP server tools display (/mcp)
1088- MCP authentication and permission improvements
1089- Added auto-reconnection for MCP SSE connections on disconnect
1090- Fixed issue where pasted content was lost when dialogs appeared
1091
1092## 1.0.17
1093
1094- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1095- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1096- MCP server list UI improvements
1097- Update Claude Code process title to display "claude" instead of "node"
1098
1099## 1.0.11
1100
1101- Claude Code can now also be used with a Claude Pro subscription
1102- Added /upgrade for smoother switching to Claude Max plans
1103- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1104- Improved shell configuration error handling
1105- Improved todo list handling during compaction
1106
1107## 1.0.10
1108
1109- Added markdown table support
1110- Improved streaming performance
1111
1112## 1.0.8
1113
1114- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1115- Increased default otel interval from 1s -> 5s
1116- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1117- Fixed a regression where search tools unnecessarily asked for permissions
1118- Added support for triggering thinking non-English languages
1119- Improved compacting UI
1120
1121## 1.0.7
1122
1123- Renamed /allowed-tools -> /permissions
1124- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1125- Deprecated claude config commands in favor of editing settings.json
1126- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1127- Improved error handling for /install-github-app
1128- Bugfixes, UI polish, and tool reliability improvements
1129
1130## 1.0.6
1131
1132- Improved edit reliability for tab-indented files
1133- Respect CLAUDE_CONFIG_DIR everywhere
1134- Reduced unnecessary tool permission prompts
1135- Added support for symlinks in @file typeahead
1136- Bugfixes, UI polish, and tool reliability improvements
1137
1138## 1.0.4
1139
1140- Fixed a bug where MCP tool errors weren't being parsed correctly
1141
1142## 1.0.1
1143
1144- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1145- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1146- Updated documentation links and OAuth process descriptions
1147
1148## 1.0.0
1149
1150- Claude Code is now generally available
1151- Introducing Sonnet 4 and Opus 4 models
1152
1153## 0.2.125
1154
1155- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1156- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1157
1158## 0.2.117
1159
1160- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1161- Introduced settings.cleanupPeriodDays
1162- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1163- Introduced --debug mode
1164
1165## 0.2.108
1166
1167- You can now send messages to Claude while it works to steer Claude in real-time
1168- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1169- Fixed a bug where thinking was not working in -p mode
1170- Fixed a regression in /cost reporting
1171- Deprecated MCP wizard interface in favor of other MCP commands
1172- Lots of other bugfixes and improvements
1173
1174## 0.2.107
1175
1176- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1177
1178## 0.2.106
1179
1180- MCP SSE server configs can now specify custom headers
1181- Fixed a bug where MCP permission prompt didn't always show correctly
1182
1183## 0.2.105
1184
1185- Claude can now search the web
1186- Moved system & account status to /status
1187- Added word movement keybindings for Vim
1188- Improved latency for startup, todo tool, and file edits
1189
1190## 0.2.102
1191
1192- Improved thinking triggering reliability
1193- Improved @mention reliability for images and folders
1194- You can now paste multiple large chunks into one prompt
1195
1196## 0.2.100
1197
1198- Fixed a crash caused by a stack overflow error
1199- Made db storage optional; missing db support disables --continue and --resume
1200
1201## 0.2.98
1202
1203- Fixed an issue where auto-compact was running twice
1204
1205## 0.2.96
1206
1207- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1208
1209## 0.2.93
1210
1211- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1212- Claude now has access to a Todo list that helps it stay on track and be more organized
1213
1214## 0.2.82
1215
1216- Added support for --disallowedTools
1217- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1218
1219## 0.2.75
1220
1221- Hit Enter to queue up additional messages while Claude is working
1222- Drag in or copy/paste image files directly into the prompt
1223- @-mention files to directly add them to context
1224- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1225- Improved performance for filename auto-complete
1226
1227## 0.2.74
1228
1229- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1230- Task tool can now perform writes and run bash commands
1231
1232## 0.2.72
1233
1234- Updated spinner to indicate tokens loaded and tool usage
1235
1236## 0.2.70
1237
1238- Network commands like curl are now available for Claude to use
1239- Claude can now run multiple web queries in parallel
1240- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1241
1242## 0.2.69
1243
1244- Fixed UI glitches with improved Select component behavior
1245- Enhanced terminal output display with better text truncation logic
1246
1247## 0.2.67
1248
1249- Shared project permission rules can be saved in .claude/settings.json
1250
1251## 0.2.66
1252
1253- Print mode (-p) now supports streaming output via --output-format=stream-json
1254- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1255
1256## 0.2.63
1257
1258- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1259
1260## 0.2.61
1261
1262- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1263- Enhanced image detection for more reliable clipboard paste functionality
1264- Fixed an issue where ESC key could crash the conversation history selector
1265
1266## 0.2.59
1267
1268- Copy+paste images directly into your prompt
1269- Improved progress indicators for bash and fetch tools
1270- Bugfixes for non-interactive mode (-p)
1271
1272## 0.2.54
1273
1274- Quickly add to Memory by starting your message with '#'
1275- Press ctrl+r to see full output for long tool results
1276- Added support for MCP SSE transport
1277
1278## 0.2.53
1279
1280- New web fetch tool lets Claude view URLs that you paste in
1281- Fixed a bug with JPEG detection
1282
1283## 0.2.50
1284
1285- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1286
1287## 0.2.49
1288
1289- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1290
1291## 0.2.47
1292
1293- Press Tab to auto-complete file and folder names
1294- Press Shift + Tab to toggle auto-accept for file edits
1295- Automatic conversation compaction for infinite conversation length (toggle with /config)
1296
1297## 0.2.44
1298
1299- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1300
1301## 0.2.41
1302
1303- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1304- MCP server startup no longer blocks the app from starting up
1305
1306## 0.2.37
1307
1308- New /release-notes command lets you view release notes at any time
1309- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1310
1311## 0.2.36
1312
1313- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1314- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1315
1316## 0.2.34
1317
1318- Vim bindings for text input - enable with /vim or /config
1319
1320## 0.2.32
1321
1322- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1323- Fix for some PersistentShell issues
1324
1325## 0.2.31
1326
1327- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1328- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1329
1330## 0.2.30
1331
1332- Added ANSI color theme for better terminal compatibility
1333- Fixed issue where slash command arguments weren't being sent properly
1334- (Mac-only) API keys are now stored in macOS Keychain
1335
1336## 0.2.26
1337
1338- New /approved-tools command for managing tool permissions
1339- Word-level diff display for improved code readability
1340- Fuzzy matching for slash commands
1341
1342## 0.2.21
1343
1344- Fuzzy matching for /commands