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 4f18698 · on Jan 9, 2026 History
1,339 lines · 63.7 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.3
4
5- Merged slash commands and skills, simplifying the mental model with no change in behavior
6- Added release channel (`stable` or `latest`) toggle to `/config`
7- 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
8- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
9- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
10- Fixed mismatch between background task count in status bar and items shown in tasks dialog
11- Fixed sub-agents using the wrong model during conversation compaction
12- Fixed web search in sub-agents using incorrect model
13- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
14- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
15- Improved slash command suggestion readability by truncating long descriptions to 2 lines
16- Changed tool hook execution timeout from 60 seconds to 10 minutes
17- [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)
18
19## 2.1.2
20
21- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
22- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
23- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
24- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
25- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
26- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
27- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
28- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
29- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
30- Fixed updates incorrectly claiming another installation is in progress
31- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
32- Fixed remote session URL and teleport being broken when using `/tasks` command
33- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
34- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
35- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
36- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
37- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
38- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
39- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
40- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
41- [SDK] Changed minimum zod peer dependency to ^4.0.0
42- [VSCode] Fixed usage display not updating after manual compact
43
44## 2.1.0
45
46- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
47- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
48- Added support for `agent` field in skills to specify agent type for execution
49- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
50- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
51- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
52- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
53- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
54- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
55- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
56- Fixed Esc key with queued prompts to only move them to input without canceling the running task
57- Reduced permission prompts for complex bash commands
58- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
59- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
60- Fixed plugin path resolution for file-based marketplace sources
61- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
62- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
63- Fixed Claude in Chrome support for WSL environments
64- Fixed Windows native installer silently failing when executable creation fails
65- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
66- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
67- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
68- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
69- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
70- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
71- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
72- Added hooks support for skill and slash command frontmatter
73- 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
74- Added `/plan` command shortcut to enable plan mode directly from the prompt
75- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
76- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
77- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
78- Added support for `once: true` config for hooks
79- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
80- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
81- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
82- Added left/right arrow key navigation for cycling through tabs in dialogs
83- Added real-time thinking block display in Ctrl+O transcript mode
84- Added filepath to full output in background bash task details dialog
85- Added Skills as a separate category in the context visualization
86- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
87- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
88- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
89- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
90- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
91- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
92- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
93- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
94- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
95- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
96- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
97- Fixed commands with `$()` command substitution failing with parse errors
98- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
99- 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)
100- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
101- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
102- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
103- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
104- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
105- Fixed subagents sometimes not inheriting the parent's model by default
106- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
107- Fixed duplicate Bash commands appearing in permission request option labels
108- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
109- Fixed background task completion notifications to appear proactively with bullet point
110- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
111- Fixed cursor disappearing after dismissing permission dialogs
112- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
113- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
114- Fixed images being silently dropped when queueing messages while backgrounding a task
115- Fixed large pasted images failing with "Image was too large" error
116- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
117- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
118- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
119- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
120- Fixed spinner token counter to properly accumulate tokens from subagents during execution
121- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
122- Fixed race condition where LSP tool could return "no server available" during startup
123- Fixed feedback submission hanging indefinitely when network requests timeout
124- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
125- Fixed hook success message showing trailing colon when hook has no output
126- Multiple optimizations to improve startup performance
127- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
128- Improved performance when reading Jupyter notebooks with many cells
129- Improved reliability for piped input like `cat refactor.md | claude`
130- Improved reliability for AskQuestion tool
131- Improved sed in-place edit commands to render as file edits with diff preview
132- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
133- Improved compaction reliability
134- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
135- Improved skills to show progress while executing, displaying tool uses as they happen
136- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
137- Improved skill suggestions to prioritize recently and frequently used skills
138- Improved spinner feedback when waiting for the first response token
139- Improved token count display in spinner to include tokens from background agents
140- Improved incremental output for async agents to give the main thread more control and visibility
141- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
142- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
143- Improved macOS screenshot paste reliability with TIFF format support
144- Improved `/stats` output
145- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
146- Changed "Interrupted" message color from red to grey for a less alarming appearance
147- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
148- Removed underline styling from image reference links
149- [SDK] Changed minimum zod peer dependency to ^4.0.0
150- [VSCode] Added currently selected model name to the context menu
151- [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")
152- [VSCode] Fixed paragraph breaks not rendering in markdown content
153- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
154- [Windows] Fixed issue with improper rendering
155
156## 2.0.76
157
158- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
159
160## 2.0.75
161
162- Minor bugfixes
163
164## 2.0.74
165
166- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
167- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
168- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
169- Added syntax highlighting info to theme picker
170- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
171- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
172- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
173- Fixed a potential crash when syntax highlighting isn't initialized correctly
174- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
175- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
176- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
177- [Windows] Fixed issue with improper rendering
178- [VSCode] Added gift tag pictogram for year-end promotion message
179
180## 2.0.73
181
182- Added clickable `[Image #N]` links that open attached images in the default viewer
183- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
184- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
185- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
186- Fixed slow input history cycling and race condition that could overwrite text after message submission
187- Improved `/theme` command to open theme picker directly
188- Improved theme picker UI
189- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
190- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
191
192## 2.0.72
193
194- 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
195- Reduced terminal flickering
196- Added scannable QR code to mobile app tip for quick app downloads
197- Added loading indicator when resuming conversations for better feedback
198- Fixed `/context` command not respecting custom system prompts in non-interactive mode
199- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
200- Improved @ mention file suggestion speed (~3x faster in git repositories)
201- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
202- Improved settings validation errors to be more prominent
203- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
204
205## 2.0.71
206
207- Added /config toggle to enable/disable prompt suggestions
208- Added `/settings` as an alias for the `/config` command
209- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
210- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
211- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
212- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
213- New syntax highlighting engine for native build
214
215## 2.0.70
216
217- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
218- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
219- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
220- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
221- Fixed input being cleared when processing queued commands while the user was typing
222- Fixed prompt suggestions replacing typed input when pressing Tab
223- Fixed diff view not updating when terminal is resized
224- Improved memory usage by 3x for large conversations
225- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
226- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
227- Fix thinking mode toggle in /config not persisting correctly
228- Improve UI for file creation permission dialog
229
230## 2.0.69
231
232- Minor bugfixes
233
234## 2.0.68
235
236- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
237- Fixed a bug where disallowed MCP tools were visible to the model
238- Fixed an issue where steering messages could be lost while a subagent is working
239- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
240- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
241- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
242
243## 2.0.67
244
245- Thinking mode is now enabled by default for Opus 4.5
246- Thinking mode configuration has moved to /config
247- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
248- Show reason why autoupdater is disabled in `/doctor`
249- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
250- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
251- Fixed scroll position resetting after deleting a permission rule in `/permissions`
252- 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
253- Fixed `claude install --force` not bypassing stale lock files
254- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
255- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
256
257## 2.0.65
258
259- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
260- Added context window information to status line input
261- Added `fileSuggestion` setting for custom `@` file search commands
262- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
263- Fixed prompt not being saved to history when aborting a query with Escape
264- Fixed Read tool image handling to identify format from bytes instead of file extension
265
266## 2.0.64
267
268- Made auto-compacting instant
269- Agents and bash commands can run asynchronously and send messages to wake up the main agent
270- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
271- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
272- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
273- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
274- Fixed auto-loading .env when using native installer
275- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
276- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
277- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
278- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
279- Bedrock: Improve efficiency of token counting
280- Bedrock: Add support for `aws login` AWS Management Console credentials
281- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
282
283## 2.0.62
284
285- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
286- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
287- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
288- Fixed slash command selection not working when multiple commands share the same name
289- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
290- Fixed running versions getting removed because lock file incorrectly going stale
291- Fixed IDE diff tab not closing when rejecting file changes
292
293## 2.0.61
294
295- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
296
297## 2.0.60
298
299- Added background agent support. Agents run in the background while you work
300- Added --disable-slash-commands CLI flag to disable all slash commands
301- Added model name to "Co-Authored-By" commit messages
302- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
303- Updated Fetch to skip summarization for pre-approved websites
304- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
305
306## 2.0.59
307
308- Added --agent CLI flag to override the agent setting for the current session
309- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
310- VS Code: Fixed .claude.json config file being read from incorrect location
311
312## 2.0.58
313
314- Pro users now have access to Opus 4.5 as part of their subscription!
315- Fixed timer duration showing "11m 60s" instead of "12m 0s"
316- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
317
318## 2.0.57
319
320- Added feedback input when rejecting plans, allowing users to tell Claude what to change
321- VSCode: Added streaming message support for real-time response display
322
323## 2.0.56
324
325- Added setting to enable/disable terminal progress bar (OSC 9;4)
326- 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.
327
328## 2.0.55
329
330- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
331- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
332- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
333- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
334
335## 2.0.54
336
337- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
338- Fix issue with excessive iTerm notifications
339
340## 2.0.52
341
342- Fixed duplicate message display when starting Claude with a command line argument
343- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
344- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
345- Permit some uses of `$!` in bash commands
346
347## 2.0.51
348
349- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
350- Introducing Claude Code for Desktop: https://claude.com/download
351- 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
352- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
353- Plan Mode now builds more precise plans and executes more thoroughly
354- Usage limit notifications now easier to understand
355- Switched `/usage` back to "% used"
356- Fixed handling of thinking errors
357- Fixed performance regression
358
359## 2.0.50
360
361- Fixed bug preventing calling MCP tools that have nested references in their input schemas
362- Silenced a noisy but harmless error during upgrades
363- Improved ultrathink text display
364- Improved clarity of 5-hour session limit warning message
365
366## 2.0.49
367
368- Added readline-style ctrl-y for pasting deleted text
369- Improved clarity of usage limit warning message
370- Fixed handling of subagent permissions
371
372## 2.0.47
373
374- Improved error messages and validation for `claude --teleport`
375- Improved error handling in `/usage`
376- Fixed race condition with history entry not getting logged at exit
377- Fixed Vertex AI configuration not being applied from `settings.json`
378
379## 2.0.46
380
381- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
382
383## 2.0.45
384
385- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
386- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
387- Send background tasks to Claude Code on the web by starting a message with `&`
388
389## 2.0.43
390
391- Added `permissionMode` field for custom agents
392- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
393- Added skills frontmatter field to declare skills to auto-load for subagents
394- Added the `SubagentStart` hook event
395- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
396- Fixed duplicate rendering of some messages in the UI
397- Fixed some visual flickers
398- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
399
400## 2.0.42
401
402- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
403
404## 2.0.41
405
406- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
407- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
408- Fixed incorrect labeling of user settings vs project settings in command descriptions
409- Fixed crash when plugin command hooks timeout during execution
410- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
411- Fixed broken security documentation links in trust dialogs and onboarding
412- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
413- ctrl-r history search landing on a slash command no longer cancels the search
414- SDK: Support custom timeouts for hooks
415- Allow more safe git commands to run without approval
416- Plugins: Added support for sharing and installing output styles
417- Teleporting a session from web will automatically set the upstream branch
418
419## 2.0.37
420
421- Fixed how idleness is computed for notifications
422- Hooks: Added matcher values for Notification hook events
423- Output Styles: Added `keep-coding-instructions` option to frontmatter
424
425## 2.0.36
426
427- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
428- Fixed queued messages being incorrectly executed as bash commands
429- Fixed input being lost when typing while a queued message is processed
430
431## 2.0.35
432
433- Improve fuzzy search results when searching commands
434- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
435- 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
436- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
437- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
438
439## 2.0.34
440
441- VSCode Extension: Added setting to configure the initial permission mode for new conversations
442- Improved file path suggestion performance with native Rust-based fuzzy finder
443- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
444- Fixed memory crash when reading or writing large files (especially base64-encoded images)
445
446## 2.0.33
447
448- Native binary installs now launch quicker.
449- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
450- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
451
452## 2.0.32
453
454- Un-deprecate output styles based on community feedback
455- Added `companyAnnouncements` setting for displaying announcements on startup
456- Fixed hook progress messages not updating correctly during PostToolUse hook execution
457
458## 2.0.31
459
460- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
461- Vertex: add support for Web Search on supported models
462- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
463- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
464- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
465- Fixed plugin uninstall not removing plugins
466
467## 2.0.30
468
469- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
470- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
471- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
472- Added prompt-based stop hooks
473- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
474- Enabled SSE MCP servers on native build
475- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
476- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
477- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
478- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
479- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
480- Fixed bug that saved session permissions to local settings
481- Fixed MCP tools not being available to sub-agents
482- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
483- Fixed delay when navigating through typeahead suggestions with arrow keys
484- VSCode: Restored selection indicator in input footer showing current file or code selection status
485
486## 2.0.28
487
488- Plan mode: introduced new Plan subagent
489- Subagents: claude can now choose to resume subagents
490- Subagents: claude can dynamically choose the model used by its subagents
491- SDK: added --max-budget-usd flag
492- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
493- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
494- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
495- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
496- Various other bug fixes
497
498## 2.0.27
499
500- New UI for permission prompts
501- Added current branch filtering and search to session resume screen for easier navigation
502- Fixed directory @-mention causing "No assistant message found" error
503- VSCode Extension: Add config setting to include .gitignored files in file searches
504- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
505
506## 2.0.25
507
508- 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
509
510## 2.0.24
511
512- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
513- Claude Code Web: Support for Web -> CLI teleport
514- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
515- Bedrock: Display awsAuthRefresh output when auth is required
516
517## 2.0.22
518
519- Fixed content layout shift when scrolling through slash commands
520- IDE: Add toggle to enable/disable thinking.
521- Fix bug causing duplicate permission prompts with parallel tool calls
522- Add support for enterprise managed MCP allowlist and denylist
523
524## 2.0.21
525
526- Support MCP `structuredContent` field in tool responses
527- Added an interactive question tool
528- Claude will now ask you questions more often in plan mode
529- Added Haiku 4.5 as a model option for Pro users
530- Fixed an issue where queued commands don't have access to previous messages' output
531
532## 2.0.20
533
534- Added support for Claude Skills
535
536## 2.0.19
537
538- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
539- Fixed a bug where Haiku was unnecessarily called in print mode
540
541## 2.0.17
542
543- Added Haiku 4.5 to model selector!
544- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
545- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
546- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
547- OTEL: support HTTP_PROXY and HTTPS_PROXY
548- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
549
550## 2.0.15
551
552- Fixed bug with resuming where previously created files needed to be read again before writing
553- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
554
555## 2.0.14
556
557- Fix @-mentioning MCP servers to toggle them on/off
558- Improve permission checks for bash with inline env vars
559- Fix ultrathink + thinking toggle
560- Reduce unnecessary logins
561- Document --system-prompt
562- Several improvements to rendering
563- Plugins UI polish
564
565## 2.0.13
566
567- Fixed `/plugin` not working on native build
568
569## 2.0.12
570
571- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
572- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
573- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
574- `/plugin validate` command for validating plugin structure and configuration
575- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
576- Plugin documentation available at https://code.claude.com/docs/en/plugins
577- Comprehensive error messages and diagnostics via `/doctor` command
578- Avoid flickering in `/model` selector
579- Improvements to `/help`
580- Avoid mentioning hooks in `/resume` summaries
581- Changes to the "verbose" setting in `/config` now persist across sessions
582
583## 2.0.11
584
585- Reduced system prompt size by 1.4k tokens
586- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
587- Fixed Opus fallback rate limit errors appearing incorrectly
588- Fixed /add-dir command selecting wrong default tab
589
590## 2.0.10
591
592- Rewrote terminal renderer for buttery smooth UI
593- Enable/disable MCP servers by @mentioning, or in /mcp
594- Added tab completion for shell commands in bash mode
595- PreToolUse hooks can now modify tool inputs
596- Press Ctrl-G to edit your prompt in your system's configured text editor
597- Fixes for bash permission checks with environment variables in the command
598
599## 2.0.9
600
601- Fix regression where bash backgrounding stopped working
602
603## 2.0.8
604
605- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
606- IDE: Add drag-and-drop support for files and folders in chat
607- /context: Fix counting for thinking blocks
608- Improve message rendering for users with light themes on dark terminals
609- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
610
611## 2.0.5
612
613- IDE: Fix IME unintended message submission with Enter and Tab
614- IDE: Add "Open in Terminal" link in login screen
615- Fix unhandled OAuth expiration 401 API errors
616- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
617
618## 2.0.1
619
620- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
621- Various bug fixes and presentation improvements
622
623## 2.0.0
624
625- New native VS Code extension
626- Fresh coat of paint throughout the whole app
627- /rewind a conversation to undo code changes
628- /usage command to see plan limits
629- Tab to toggle thinking (sticky across sessions)
630- Ctrl-R to search history
631- Unshipped claude config command
632- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
633- SDK: The Claude Code SDK is now the Claude Agent SDK
634- Add subagents dynamically with `--agents` flag
635
636## 1.0.126
637
638- Enable /context command for Bedrock and Vertex
639- Add mTLS support for HTTP-based OpenTelemetry exporters
640
641## 1.0.124
642
643- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
644- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
645- No longer inform Claude of the list of allowed tools when permission is denied
646- Fixed security vulnerability in Bash tool permission checks
647- Improved VSCode extension performance for large files
648
649## 1.0.123
650
651- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
652- Fixed thinking mode triggering on negation phrases like "don't think"
653- Fixed rendering performance degradation during token streaming
654- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
655- Enhanced BashTool environment snapshot logging
656- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
657- Migrated --debug logging to a file, to enable easy tailing & filtering
658
659## 1.0.120
660
661- Fix input lag during typing, especially noticeable with large prompts
662- Improved VSCode extension command registry and sessions dialog user experience
663- Enhanced sessions dialog responsiveness and visual feedback
664- Fixed IDE compatibility issue by removing worktree support check
665- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
666
667## 1.0.119
668
669- Fix Windows issue where process visually freezes on entering interactive mode
670- Support dynamic headers for MCP servers via headersHelper configuration
671- Fix thinking mode not working in headless sessions
672- Fix slash commands now properly update allowed tools instead of replacing them
673
674## 1.0.117
675
676- Add Ctrl-R history search to recall previous commands like bash/zsh
677- Fix input lag while typing, especially on Windows
678- Add sed command to auto-allowed commands in acceptEdits mode
679- Fix Windows PATH comparison to be case-insensitive for drive letters
680- Add permissions management hint to /add-dir output
681
682## 1.0.115
683
684- Improve thinking mode display with enhanced visual effects
685- Type /t to temporarily disable thinking mode in your prompt
686- Improve path validation for glob and grep tools
687- Show condensed output for post-tool hooks to reduce visual clutter
688- Fix visual feedback when loading state completes
689- Improve UI consistency for permission request dialogs
690
691## 1.0.113
692
693- Deprecated piped input in interactive mode
694- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
695
696## 1.0.112
697
698- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
699- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
700- Hooks: Added systemMessage support for SessionEnd hooks
701- Added `spinnerTipsEnabled` setting to disable spinner tips
702- IDE: Various improvements and bug fixes
703
704## 1.0.111
705
706- /model now validates provided model names
707- Fixed Bash tool crashes caused by malformed shell syntax parsing
708
709## 1.0.110
710
711- /terminal-setup command now supports WezTerm
712- MCP: OAuth tokens now proactively refresh before expiration
713- Fixed reliability issues with background Bash processes
714
715## 1.0.109
716
717- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
718
719## 1.0.106
720
721- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
722
723## 1.0.97
724
725- Settings: /doctor now validates permission rule syntax and suggests corrections
726
727## 1.0.94
728
729- Vertex: add support for global endpoints for supported models
730- /memory command now allows direct editing of all imported memory files
731- SDK: Add custom tools as callbacks
732- Added /todos command to list current todo items
733
734## 1.0.93
735
736- Windows: Add alt + v shortcut for pasting images from clipboard
737- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
738
739## 1.0.90
740
741- Settings file changes take effect immediately - no restart required
742
743## 1.0.88
744
745- Fixed issue causing "OAuth authentication is currently not supported"
746- Status line input now includes `exceeds_200k_tokens`
747- Fixed incorrect usage tracking in /cost.
748- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
749- Bedrock: Updated default Sonnet model to Sonnet 4
750
751## 1.0.86
752
753- Added /context to help users self-serve debug context issues
754- SDK: Added UUID support for all SDK messages
755- SDK: Added `--replay-user-messages` to replay user messages back to stdout
756
757## 1.0.85
758
759- Status line input now includes session cost info
760- Hooks: Introduced SessionEnd hook
761
762## 1.0.84
763
764- Fix tool_use/tool_result id mismatch error when network is unstable
765- Fix Claude sometimes ignoring real-time steering when wrapping up a task
766- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
767- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
768
769## 1.0.83
770
771- @-mention: Support files with spaces in path
772- New shimmering spinner
773
774## 1.0.82
775
776- SDK: Add request cancellation support
777- SDK: New additionalDirectories option to search custom paths, improved slash command processing
778- Settings: Validation prevents invalid fields in .claude/settings.json files
779- MCP: Improve tool name consistency
780- Bash: Fix crash when Claude tries to automatically read large files
781
782## 1.0.81
783
784- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
785- Agents: Fix custom agent loading when agent files are unparsable
786
787## 1.0.80
788
789- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
790
791## 1.0.77
792
793- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
794- SDK: Add session support and permission denial tracking
795- Fix token limit errors in conversation summarization
796- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
797
798## 1.0.73
799
800- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
801- MCP: Press Esc to cancel OAuth authentication flows
802- Bash: Improved command validation and reduced false security warnings
803- UI: Enhanced spinner animations and status line visual hierarchy
804- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
805
806## 1.0.72
807
808- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
809
810## 1.0.71
811
812- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
813- Customizable status line: add your terminal prompt to Claude Code with /statusline
814
815## 1.0.70
816
817- Performance: Optimized message rendering for better performance with large contexts
818- Windows: Fixed native file search, ripgrep, and subagent functionality
819- Added support for @-mentions in slash command arguments
820
821## 1.0.69
822
823- Upgraded Opus to version 4.1
824
825## 1.0.68
826
827- Fix incorrect model names being used for certain commands like `/pr-comments`
828- 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.
829- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
830- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
831- SDK: Added canUseTool callback support for tool confirmation
832- Added `disableAllHooks` setting
833- Improved file suggestions performance in large repos
834
835## 1.0.65
836
837- IDE: Fixed connection stability issues and error handling for diagnostics
838- Windows: Fixed shell environment setup for users without .bashrc files
839
840## 1.0.64
841
842- Agents: Added model customization support - you can now specify which model an agent should use
843- Agents: Fixed unintended access to the recursive agent tool
844- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
845- SDK: Fixed user input tracking across multi-turn conversations
846- Added hidden files to file search and @-mention suggestions
847
848## 1.0.63
849
850- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
851
852## 1.0.62
853
854- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
855- Hooks: Added SessionStart hook for new session initialization
856- /add-dir command now supports typeahead for directory paths
857- Improved network connectivity check reliability
858
859## 1.0.61
860
861- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
862- Settings: Added `--settings` flag to load settings from a JSON file
863- Settings: Fixed resolution of settings files paths that are symlinks
864- OTEL: Fixed reporting of wrong organization after authentication changes
865- Slash commands: Fixed permissions checking for allowed-tools with Bash
866- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
867- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
868- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
869
870## 1.0.60
871
872- You can now create custom subagents for specialized tasks! Run /agents to get started
873
874## 1.0.59
875
876- SDK: Added tool confirmation support with canUseTool callback
877- SDK: Allow specifying env for spawned process
878- Hooks: Exposed PermissionDecision to hooks (including "ask")
879- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
880- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
881
882## 1.0.58
883
884- Added support for reading PDFs
885- MCP: Improved server health status display in 'claude mcp list'
886- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
887
888## 1.0.57
889
890- Added support for specifying a model in slash commands
891- Improved permission messages to help Claude understand allowed tools
892- Fix: Remove trailing newlines from bash output in terminal wrapping
893
894## 1.0.56
895
896- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
897- Fixes for WSL IDE detection
898- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
899
900## 1.0.55
901
902- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
903- Windows: fixed Ctrl+Z crash
904- SDK: Added ability to capture error logging
905- Add --system-prompt-file option to override system prompt in print mode
906
907## 1.0.54
908
909- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
910- Custom slash commands: Added argument-hint to frontmatter
911- Windows: OAuth uses port 45454 and properly constructs browser URL
912- Windows: mode switching now uses alt + m, and plan mode renders properly
913- Shell: Switch to in-memory shell snapshot to fix file-related errors
914
915## 1.0.53
916
917- Updated @-mention file truncation from 100 lines to 2000 lines
918- 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).
919
920## 1.0.52
921
922- Added support for MCP server instructions
923
924## 1.0.51
925
926- Added support for native Windows (requires Git for Windows)
927- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
928- Settings: /doctor can now help you identify and fix invalid setting files
929- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
930- Increased auto-compact warning threshold from 60% to 80%
931- Fixed an issue with handling user directories with spaces for shell snapshots
932- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
933- Custom slash commands: Fixed user-level commands in subdirectories
934- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
935
936## 1.0.48
937
938- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
939- Added progress messages to Bash tool based on the last 5 lines of command output
940- Added expanding variables support for MCP server configuration
941- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
942- Improved IDE extension path handling when Claude Code runs in WSL
943- Hooks: Added a PreCompact hook
944- Vim mode: Added c, f/F, t/T
945
946## 1.0.45
947
948- Redesigned Search (Grep) tool with new tool input parameters and features
949- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
950- Fixed config file corruption issue by enforcing atomic writes
951- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
952- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
953- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
954
955## 1.0.44
956
957- New /export command lets you quickly export a conversation for sharing
958- MCP: resource_link tool results are now supported
959- MCP: tool annotations and tool titles now display in /mcp view
960- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
961
962## 1.0.43
963
964- Fixed a bug where the theme selector was saving excessively
965- Hooks: Added EPIPE system error handling
966
967## 1.0.42
968
969- Added tilde (`~`) expansion support to `/add-dir` command
970
971## 1.0.41
972
973- Hooks: Split Stop hook triggering into Stop and SubagentStop
974- Hooks: Enabled optional timeout configuration for each command
975- Hooks: Added "hook_event_name" to hook input
976- Fixed a bug where MCP tools would display twice in tool list
977- New tool parameters JSON for Bash tool in `tool_decision` event
978
979## 1.0.40
980
981- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
982
983## 1.0.39
984
985- New Active Time metric in OpenTelemetry logging
986
987## 1.0.38
988
989- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
990
991## 1.0.37
992
993- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
994
995## 1.0.36
996
997- Web search now takes today's date into context
998- Fixed a bug where stdio MCP servers were not terminating properly on exit
999
1000## 1.0.35
1001
1002- Added support for MCP OAuth Authorization Server discovery
1003
1004## 1.0.34
1005
1006- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1007
1008## 1.0.33
1009
1010- Improved logging functionality with session ID support
1011- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1012- Improvements to plan mode
1013
1014## 1.0.32
1015
1016- Updated loopback config for litellm
1017- Added forceLoginMethod setting to bypass login selection screen
1018
1019## 1.0.31
1020
1021- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1022
1023## 1.0.30
1024
1025- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1026- Improved file path autocomplete with filename matching
1027- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1028- Enhanced jq regex support for complex filters with pipes and select
1029
1030## 1.0.29
1031
1032- Improved CJK character support in cursor navigation and rendering
1033
1034## 1.0.28
1035
1036- Slash commands: Fix selector display during history navigation
1037- Resizes images before upload to prevent API size limit errors
1038- Added XDG_CONFIG_HOME support to configuration directory
1039- Performance optimizations for memory usage
1040- New attributes (terminal.type, language) in OpenTelemetry logging
1041
1042## 1.0.27
1043
1044- Streamable HTTP MCP servers are now supported
1045- Remote MCP servers (SSE and HTTP) now support OAuth
1046- MCP resources can now be @-mentioned
1047- /resume slash command to switch conversations within Claude Code
1048
1049## 1.0.25
1050
1051- Slash commands: moved "project" and "user" prefixes to descriptions
1052- Slash commands: improved reliability for command discovery
1053- Improved support for Ghostty
1054- Improved web search reliability
1055
1056## 1.0.24
1057
1058- Improved /mcp output
1059- Fixed a bug where settings arrays got overwritten instead of merged
1060
1061## 1.0.23
1062
1063- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1064- Released Python SDK: pip install claude-code-sdk to get started
1065
1066## 1.0.22
1067
1068- SDK: Renamed `total_cost` to `total_cost_usd`
1069
1070## 1.0.21
1071
1072- Improved editing of files with tab-based indentation
1073- Fix for tool_use without matching tool_result errors
1074- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1075
1076## 1.0.18
1077
1078- Added --add-dir CLI argument for specifying additional working directories
1079- Added streaming input support without require -p flag
1080- Improved startup performance and session storage performance
1081- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1082- Added detailed MCP server tools display (/mcp)
1083- MCP authentication and permission improvements
1084- Added auto-reconnection for MCP SSE connections on disconnect
1085- Fixed issue where pasted content was lost when dialogs appeared
1086
1087## 1.0.17
1088
1089- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1090- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1091- MCP server list UI improvements
1092- Update Claude Code process title to display "claude" instead of "node"
1093
1094## 1.0.11
1095
1096- Claude Code can now also be used with a Claude Pro subscription
1097- Added /upgrade for smoother switching to Claude Max plans
1098- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1099- Improved shell configuration error handling
1100- Improved todo list handling during compaction
1101
1102## 1.0.10
1103
1104- Added markdown table support
1105- Improved streaming performance
1106
1107## 1.0.8
1108
1109- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1110- Increased default otel interval from 1s -> 5s
1111- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1112- Fixed a regression where search tools unnecessarily asked for permissions
1113- Added support for triggering thinking non-English languages
1114- Improved compacting UI
1115
1116## 1.0.7
1117
1118- Renamed /allowed-tools -> /permissions
1119- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1120- Deprecated claude config commands in favor of editing settings.json
1121- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1122- Improved error handling for /install-github-app
1123- Bugfixes, UI polish, and tool reliability improvements
1124
1125## 1.0.6
1126
1127- Improved edit reliability for tab-indented files
1128- Respect CLAUDE_CONFIG_DIR everywhere
1129- Reduced unnecessary tool permission prompts
1130- Added support for symlinks in @file typeahead
1131- Bugfixes, UI polish, and tool reliability improvements
1132
1133## 1.0.4
1134
1135- Fixed a bug where MCP tool errors weren't being parsed correctly
1136
1137## 1.0.1
1138
1139- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1140- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1141- Updated documentation links and OAuth process descriptions
1142
1143## 1.0.0
1144
1145- Claude Code is now generally available
1146- Introducing Sonnet 4 and Opus 4 models
1147
1148## 0.2.125
1149
1150- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1151- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1152
1153## 0.2.117
1154
1155- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1156- Introduced settings.cleanupPeriodDays
1157- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1158- Introduced --debug mode
1159
1160## 0.2.108
1161
1162- You can now send messages to Claude while it works to steer Claude in real-time
1163- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1164- Fixed a bug where thinking was not working in -p mode
1165- Fixed a regression in /cost reporting
1166- Deprecated MCP wizard interface in favor of other MCP commands
1167- Lots of other bugfixes and improvements
1168
1169## 0.2.107
1170
1171- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1172
1173## 0.2.106
1174
1175- MCP SSE server configs can now specify custom headers
1176- Fixed a bug where MCP permission prompt didn't always show correctly
1177
1178## 0.2.105
1179
1180- Claude can now search the web
1181- Moved system & account status to /status
1182- Added word movement keybindings for Vim
1183- Improved latency for startup, todo tool, and file edits
1184
1185## 0.2.102
1186
1187- Improved thinking triggering reliability
1188- Improved @mention reliability for images and folders
1189- You can now paste multiple large chunks into one prompt
1190
1191## 0.2.100
1192
1193- Fixed a crash caused by a stack overflow error
1194- Made db storage optional; missing db support disables --continue and --resume
1195
1196## 0.2.98
1197
1198- Fixed an issue where auto-compact was running twice
1199
1200## 0.2.96
1201
1202- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1203
1204## 0.2.93
1205
1206- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1207- Claude now has access to a Todo list that helps it stay on track and be more organized
1208
1209## 0.2.82
1210
1211- Added support for --disallowedTools
1212- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1213
1214## 0.2.75
1215
1216- Hit Enter to queue up additional messages while Claude is working
1217- Drag in or copy/paste image files directly into the prompt
1218- @-mention files to directly add them to context
1219- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1220- Improved performance for filename auto-complete
1221
1222## 0.2.74
1223
1224- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1225- Task tool can now perform writes and run bash commands
1226
1227## 0.2.72
1228
1229- Updated spinner to indicate tokens loaded and tool usage
1230
1231## 0.2.70
1232
1233- Network commands like curl are now available for Claude to use
1234- Claude can now run multiple web queries in parallel
1235- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1236
1237## 0.2.69
1238
1239- Fixed UI glitches with improved Select component behavior
1240- Enhanced terminal output display with better text truncation logic
1241
1242## 0.2.67
1243
1244- Shared project permission rules can be saved in .claude/settings.json
1245
1246## 0.2.66
1247
1248- Print mode (-p) now supports streaming output via --output-format=stream-json
1249- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1250
1251## 0.2.63
1252
1253- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1254
1255## 0.2.61
1256
1257- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1258- Enhanced image detection for more reliable clipboard paste functionality
1259- Fixed an issue where ESC key could crash the conversation history selector
1260
1261## 0.2.59
1262
1263- Copy+paste images directly into your prompt
1264- Improved progress indicators for bash and fetch tools
1265- Bugfixes for non-interactive mode (-p)
1266
1267## 0.2.54
1268
1269- Quickly add to Memory by starting your message with '#'
1270- Press ctrl+r to see full output for long tool results
1271- Added support for MCP SSE transport
1272
1273## 0.2.53
1274
1275- New web fetch tool lets Claude view URLs that you paste in
1276- Fixed a bug with JPEG detection
1277
1278## 0.2.50
1279
1280- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1281
1282## 0.2.49
1283
1284- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1285
1286## 0.2.47
1287
1288- Press Tab to auto-complete file and folder names
1289- Press Shift + Tab to toggle auto-accept for file edits
1290- Automatic conversation compaction for infinite conversation length (toggle with /config)
1291
1292## 0.2.44
1293
1294- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1295
1296## 0.2.41
1297
1298- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1299- MCP server startup no longer blocks the app from starting up
1300
1301## 0.2.37
1302
1303- New /release-notes command lets you view release notes at any time
1304- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1305
1306## 0.2.36
1307
1308- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1309- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1310
1311## 0.2.34
1312
1313- Vim bindings for text input - enable with /vim or /config
1314
1315## 0.2.32
1316
1317- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1318- Fix for some PersistentShell issues
1319
1320## 0.2.31
1321
1322- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1323- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1324
1325## 0.2.30
1326
1327- Added ANSI color theme for better terminal compatibility
1328- Fixed issue where slash command arguments weren't being sent properly
1329- (Mac-only) API keys are now stored in macOS Keychain
1330
1331## 0.2.26
1332
1333- New /approved-tools command for managing tool permissions
1334- Word-level diff display for improved code readability
1335- Fuzzy matching for slash commands
1336
1337## 0.2.21
1338
1339- Fuzzy matching for /commands