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 5862adf · on Jan 30, 2026 History
1,565 lines · 80.1 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.27
4
5- Added tool call failures and denials to debug logs
6- Fixed context management validation error for gateway users, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
7- Added `--from-pr` flag to resume sessions linked to a specific GitHub PR number or URL
8- Sessions are now automatically linked to PRs when created via `gh pr create`
9- Fixed /context command not displaying colored output
10- Fixed status bar duplicating background task indicator when PR status was shown
11- VSCode: Enabled Claude in Chrome integration
12- Permissions now respect content-level `ask` over tool-level `allow`. Previously `allow: ["Bash"], ask: ["Bash(rm *)"]` allowed all bash commands, but will now permission prompt for `rm`.
13- Windows: Fixed bash command execution failing for users with `.bashrc` files
14- Windows: Fixed console windows flashing when spawning child processes
15- VSCode: Fixed OAuth token expiration causing 401 errors after extended sessions
16
17## 2.1.25
18
19- Fixed beta header validation error for gateway users on Bedrock and Vertex, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
20
21## 2.1.23
22
23- Added customizable spinner verbs setting (`spinnerVerbs`)
24- Fixed mTLS and proxy connectivity for users behind corporate proxies or using client certificates
25- Fixed per-user temp directory isolation to prevent permission conflicts on shared systems
26- Fixed a race condition that could cause 400 errors when prompt caching scope was enabled
27- Fixed pending async hooks not being cancelled when headless streaming sessions ended
28- Fixed tab completion not updating the input field when accepting a suggestion
29- Fixed ripgrep search timeouts silently returning empty results instead of reporting errors
30- Improved terminal rendering performance with optimized screen data layout
31- Changed Bash commands to show timeout duration alongside elapsed time
32- Changed merged pull requests to show a purple status indicator in the prompt footer
33- [IDE] Fixed model options displaying incorrect region strings for Bedrock users in headless mode
34
35## 2.1.22
36
37- Fixed structured outputs for non-interactive (-p) mode
38
39## 2.1.21
40
41- Added support for full-width (zenkaku) number input from Japanese IME in option selection prompts
42- Fixed shell completion cache files being truncated on exit
43- Fixed API errors when resuming sessions that were interrupted during tool execution
44- Fixed auto-compact triggering too early on models with large output token limits
45- Fixed task IDs potentially being reused after deletion
46- Fixed file search not working in VS Code extension on Windows
47- Improved read/search progress indicators to show "Reading…" while in progress and "Read" when complete
48- Improved Claude to prefer file operation tools (Read, Edit, Write) over bash equivalents (cat, sed, awk)
49- [VSCode] Added automatic Python virtual environment activation, ensuring `python` and `pip` commands use the correct interpreter (configurable via `claudeCode.usePythonEnvironment` setting)
50- [VSCode] Fixed message action buttons having incorrect background colors
51
52## 2.1.20
53
54- Added arrow key history navigation in vim normal mode when cursor cannot move further
55- Added external editor shortcut (Ctrl+G) to the help menu for better discoverability
56- Added PR review status indicator to the prompt footer, showing the current branch's PR state (approved, changes requested, pending, or draft) as a colored dot with a clickable link
57- Added support for loading `CLAUDE.md` files from additional directories specified via `--add-dir` flag (requires setting `CLAUDE_CODE_ADDITIONAL_DIRECTORIES_CLAUDE_MD=1`)
58- Added ability to delete tasks via the `TaskUpdate` tool
59- Fixed session compaction issues that could cause resume to load full history instead of the compact summary
60- Fixed agents sometimes ignoring user messages sent while actively working on a task
61- Fixed wide character (emoji, CJK) rendering artifacts where trailing columns were not cleared when replaced by narrower characters
62- Fixed JSON parsing errors when MCP tool responses contain special Unicode characters
63- Fixed up/down arrow keys in multi-line and wrapped text input to prioritize cursor movement over history navigation
64- Fixed draft prompt being lost when pressing UP arrow to navigate command history
65- Fixed ghost text flickering when typing slash commands mid-input
66- Fixed marketplace source removal not properly deleting settings
67- Fixed duplicate output in some commands like `/context`
68- Fixed task list sometimes showing outside the main conversation view
69- Fixed syntax highlighting for diffs occurring within multiline constructs like Python docstrings
70- Fixed crashes when cancelling tool use
71- Improved `/sandbox` command UI to show dependency status with installation instructions when dependencies are missing
72- Improved thinking status text with a subtle shimmer animation
73- Improved task list to dynamically adjust visible items based on terminal height
74- Improved fork conversation hint to show how to resume the original session
75- Changed collapsed read/search groups to show present tense ("Reading", "Searching for") while in progress, and past tense ("Read", "Searched for") when complete
76- Changed `ToolSearch` results to appear as a brief notification instead of inline in the conversation
77- Changed the `/commit-push-pr` skill to automatically post PR URLs to Slack channels when configured via MCP tools
78- Changed the `/copy` command to be available to all users
79- Changed background agents to prompt for tool permissions before launching
80- Changed permission rules like `Bash(*)` to be accepted and treated as equivalent to `Bash`
81- Changed config backups to be timestamped and rotated (keeping 5 most recent) to prevent data loss
82
83## 2.1.19
84
85- Added env var `CLAUDE_CODE_ENABLE_TASKS`, set to `false` to keep the old system temporarily
86- Added shorthand `$0`, `$1`, etc. for accessing individual arguments in custom commands
87- Fixed crashes on processors without AVX instruction support
88- Fixed dangling Claude Code processes when terminal is closed by catching EIO errors from `process.exit()` and using SIGKILL as fallback
89- Fixed `/rename` and `/tag` not updating the correct session when resuming from a different directory (e.g., git worktrees)
90- Fixed resuming sessions by custom title not working when run from a different directory
91- Fixed pasted text content being lost when using prompt stash (Ctrl+S) and restore
92- Fixed agent list displaying "Sonnet (default)" instead of "Inherit (default)" for agents without an explicit model setting
93- Fixed backgrounded hook commands not returning early, potentially causing the session to wait on a process that was intentionally backgrounded
94- Fixed file write preview omitting empty lines
95- Changed skills without additional permissions or hooks to be allowed without requiring approval
96- Changed indexed argument syntax from `$ARGUMENTS.0` to `$ARGUMENTS[0]` (bracket syntax)
97- [SDK] Added replay of `queued_command` attachment messages as `SDKUserMessageReplay` events when `replayUserMessages` is enabled
98- [VSCode] Enabled session forking and rewind functionality for all users
99
100## 2.1.18
101
102- Added customizable keyboard shortcuts. Configure keybindings per context, create chord sequences, and personalize your workflow. Run `/keybindings` to get started. Learn more at https://code.claude.com/docs/en/keybindings
103
104## 2.1.17
105
106- Fixed crashes on processors without AVX instruction support
107
108## 2.1.16
109
110- Added new task management system, including new capabilities like dependency tracking
111- [VSCode] Added native plugin management support
112- [VSCode] Added ability for OAuth users to browse and resume remote Claude sessions from the Sessions dialog
113- Fixed out-of-memory crashes when resuming sessions with heavy subagent usage
114- Fixed an issue where the "context remaining" warning was not hidden after running `/compact`
115- Fixed session titles on the resume screen not respecting the user's language setting
116- [IDE] Fixed a race condition on Windows where the Claude Code sidebar view container would not appear on start
117
118## 2.1.15
119
120- Added deprecation notification for npm installations - run `claude install` or see https://docs.anthropic.com/en/docs/claude-code/getting-started for more options
121- Improved UI rendering performance with React Compiler
122- Fixed the "Context left until auto-compact" warning not disappearing after running `/compact`
123- Fixed MCP stdio server timeout not killing child process, which could cause UI freezes
124
125## 2.1.14
126
127- Added history-based autocomplete in bash mode (`!`) - type a partial command and press Tab to complete from your bash command history
128- Added search to installed plugins list - type to filter by name or description
129- Added support for pinning plugins to specific git commit SHAs, allowing marketplace entries to install exact versions
130- Fixed a regression where the context window blocking limit was calculated too aggressively, blocking users at ~65% context usage instead of the intended ~98%
131- Fixed memory issues that could cause crashes when running parallel subagents
132- Fixed memory leak in long-running sessions where stream resources were not cleaned up after shell commands completed
133- Fixed `@` symbol incorrectly triggering file autocomplete suggestions in bash mode
134- Fixed `@`-mention menu folder click behavior to navigate into directories instead of selecting them
135- Fixed `/feedback` command generating invalid GitHub issue URLs when description is very long
136- Fixed `/context` command to show the same token count and percentage as the status line in verbose mode
137- Fixed an issue where `/config`, `/context`, `/model`, and `/todos` command overlays could close unexpectedly
138- Fixed slash command autocomplete selecting wrong command when typing similar commands (e.g., `/context` vs `/compact`)
139- Fixed inconsistent back navigation in plugin marketplace when only one marketplace is configured
140- Fixed iTerm2 progress bar not clearing properly on exit, preventing lingering indicators and bell sounds
141- Improved backspace to delete pasted text as a single token instead of one character at a time
142- [VSCode] Added `/usage` command to display current plan usage
143
144## 2.1.12
145
146- Fixed message rendering bug
147
148## 2.1.11
149
150- Fixed excessive MCP connection requests for HTTP/SSE transports
151
152## 2.1.10
153
154- Added new `Setup` hook event that can be triggered via `--init`, `--init-only`, or `--maintenance` CLI flags for repository setup and maintenance operations
155- Added keyboard shortcut 'c' to copy OAuth URL when browser doesn't open automatically during login
156- Fixed a crash when running bash commands containing heredocs with JavaScript template literals like `${index + 1}`
157- Improved startup to capture keystrokes typed before the REPL is fully ready
158- Improved file suggestions to show as removable attachments instead of inserting text when accepted
159- [VSCode] Added install count display to plugin listings
160- [VSCode] Added trust warning when installing plugins
161
162## 2.1.9
163
164- Added `auto:N` syntax for configuring the MCP tool search auto-enable threshold, where N is the context window percentage (0-100)
165- Added `plansDirectory` setting to customize where plan files are stored
166- Added external editor support (Ctrl+G) in AskUserQuestion "Other" input field
167- Added session URL attribution to commits and PRs created from web sessions
168- Added support for `PreToolUse` hooks to return `additionalContext` to the model
169- Added `${CLAUDE_SESSION_ID}` string substitution for skills to access the current session ID
170- Fixed long sessions with parallel tool calls failing with an API error about orphan tool_result blocks
171- Fixed MCP server reconnection hanging when cached connection promise never resolves
172- Fixed Ctrl+Z suspend not working in terminals using Kitty keyboard protocol (Ghostty, iTerm2, kitty, WezTerm)
173
174## 2.1.7
175
176- Added `showTurnDuration` setting to hide turn duration messages (e.g., "Cooked for 1m 6s")
177- Added ability to provide feedback when accepting permission prompts
178- Added inline display of agent's final response in task notifications, making it easier to see results without reading the full transcript file
179- Fixed security vulnerability where wildcard permission rules could match compound commands containing shell operators
180- Fixed false "file modified" errors on Windows when cloud sync tools, antivirus scanners, or Git touch file timestamps without changing content
181- Fixed orphaned tool_result errors when sibling tools fail during streaming execution
182- Fixed context window blocking limit being calculated using the full context window instead of the effective context window (which reserves space for max output tokens)
183- Fixed spinner briefly flashing when running local slash commands like `/model` or `/theme`
184- Fixed terminal title animation jitter by using fixed-width braille characters
185- Fixed plugins with git submodules not being fully initialized when installed
186- Fixed bash commands failing on Windows when temp directory paths contained characters like `t` or `n` that were misinterpreted as escape sequences
187- Improved typing responsiveness by reducing memory allocation overhead in terminal rendering
188- Enabled MCP tool search auto mode by default for all users. When MCP tool descriptions exceed 10% of the context window, they are automatically deferred and discovered via the MCPSearch tool instead of being loaded upfront. This reduces context usage for users with many MCP tools configured. Users can disable this by adding `MCPSearch` to `disallowedTools` in their settings.
189- Changed OAuth and API Console URLs from console.anthropic.com to platform.claude.com
190- [VSCode] Fixed `claudeProcessWrapper` setting passing the wrapper path instead of the Claude binary path
191
192## 2.1.6
193
194- Added search functionality to `/config` command for quickly filtering settings
195- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
196- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
197- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
198- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
199- Added an error display when the editor fails during Ctrl+G
200- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
201- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
202- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
203- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
204- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
205- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
206- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
207- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
208- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
209- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
210- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
211- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
212- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
213- Improved @ autocomplete with icons for different suggestion types and single-line formatting
214- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
215- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
216- Changed terminal title to "Claude Code" on startup for better window identification
217- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
218- [VSCode] Fixed usage indicator not updating after manual compact
219
220## 2.1.5
221
222- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
223
224## 2.1.4
225
226- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
227- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
228
229## 2.1.3
230
231- Merged slash commands and skills, simplifying the mental model with no change in behavior
232- Added release channel (`stable` or `latest`) toggle to `/config`
233- 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
234- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
235- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
236- Fixed mismatch between background task count in status bar and items shown in tasks dialog
237- Fixed sub-agents using the wrong model during conversation compaction
238- Fixed web search in sub-agents using incorrect model
239- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
240- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
241- Improved slash command suggestion readability by truncating long descriptions to 2 lines
242- Changed tool hook execution timeout from 60 seconds to 10 minutes
243- [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)
244
245## 2.1.2
246
247- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
248- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
249- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
250- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
251- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
252- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
253- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
254- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
255- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
256- Fixed updates incorrectly claiming another installation is in progress
257- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
258- Fixed remote session URL and teleport being broken when using `/tasks` command
259- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
260- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
261- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
262- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
263- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
264- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
265- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
266- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
267- [SDK] Changed minimum zod peer dependency to ^4.0.0
268- [VSCode] Fixed usage display not updating after manual compact
269
270## 2.1.0
271
272- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
273- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
274- Added support for `agent` field in skills to specify agent type for execution
275- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
276- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
277- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
278- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
279- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
280- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
281- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
282- Fixed Esc key with queued prompts to only move them to input without canceling the running task
283- Reduced permission prompts for complex bash commands
284- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
285- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
286- Fixed plugin path resolution for file-based marketplace sources
287- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
288- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
289- Fixed Claude in Chrome support for WSL environments
290- Fixed Windows native installer silently failing when executable creation fails
291- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
292- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
293- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
294- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
295- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
296- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
297- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
298- Added hooks support for skill and slash command frontmatter
299- 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
300- Added `/plan` command shortcut to enable plan mode directly from the prompt
301- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
302- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
303- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
304- Added support for `once: true` config for hooks
305- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
306- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
307- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
308- Added left/right arrow key navigation for cycling through tabs in dialogs
309- Added real-time thinking block display in Ctrl+O transcript mode
310- Added filepath to full output in background bash task details dialog
311- Added Skills as a separate category in the context visualization
312- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
313- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
314- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
315- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
316- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
317- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
318- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
319- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
320- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
321- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
322- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
323- Fixed commands with `$()` command substitution failing with parse errors
324- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
325- 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)
326- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
327- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
328- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
329- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
330- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
331- Fixed subagents sometimes not inheriting the parent's model by default
332- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
333- Fixed duplicate Bash commands appearing in permission request option labels
334- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
335- Fixed background task completion notifications to appear proactively with bullet point
336- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
337- Fixed cursor disappearing after dismissing permission dialogs
338- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
339- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
340- Fixed images being silently dropped when queueing messages while backgrounding a task
341- Fixed large pasted images failing with "Image was too large" error
342- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
343- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
344- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
345- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
346- Fixed spinner token counter to properly accumulate tokens from subagents during execution
347- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
348- Fixed race condition where LSP tool could return "no server available" during startup
349- Fixed feedback submission hanging indefinitely when network requests timeout
350- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
351- Fixed hook success message showing trailing colon when hook has no output
352- Multiple optimizations to improve startup performance
353- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
354- Improved performance when reading Jupyter notebooks with many cells
355- Improved reliability for piped input like `cat refactor.md | claude`
356- Improved reliability for AskQuestion tool
357- Improved sed in-place edit commands to render as file edits with diff preview
358- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
359- Improved compaction reliability
360- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
361- Improved skills to show progress while executing, displaying tool uses as they happen
362- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
363- Improved skill suggestions to prioritize recently and frequently used skills
364- Improved spinner feedback when waiting for the first response token
365- Improved token count display in spinner to include tokens from background agents
366- Improved incremental output for async agents to give the main thread more control and visibility
367- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
368- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
369- Improved macOS screenshot paste reliability with TIFF format support
370- Improved `/stats` output
371- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
372- Changed "Interrupted" message color from red to grey for a less alarming appearance
373- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
374- Removed underline styling from image reference links
375- [SDK] Changed minimum zod peer dependency to ^4.0.0
376- [VSCode] Added currently selected model name to the context menu
377- [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")
378- [VSCode] Fixed paragraph breaks not rendering in markdown content
379- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
380- [Windows] Fixed issue with improper rendering
381
382## 2.0.76
383
384- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
385
386## 2.0.75
387
388- Minor bugfixes
389
390## 2.0.74
391
392- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
393- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
394- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
395- Added syntax highlighting info to theme picker
396- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
397- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
398- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
399- Fixed a potential crash when syntax highlighting isn't initialized correctly
400- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
401- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
402- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
403- [Windows] Fixed issue with improper rendering
404- [VSCode] Added gift tag pictogram for year-end promotion message
405
406## 2.0.73
407
408- Added clickable `[Image #N]` links that open attached images in the default viewer
409- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
410- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
411- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
412- Fixed slow input history cycling and race condition that could overwrite text after message submission
413- Improved `/theme` command to open theme picker directly
414- Improved theme picker UI
415- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
416- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
417
418## 2.0.72
419
420- 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
421- Reduced terminal flickering
422- Added scannable QR code to mobile app tip for quick app downloads
423- Added loading indicator when resuming conversations for better feedback
424- Fixed `/context` command not respecting custom system prompts in non-interactive mode
425- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
426- Improved @ mention file suggestion speed (~3× faster in git repositories)
427- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
428- Improved settings validation errors to be more prominent
429- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
430
431## 2.0.71
432
433- Added /config toggle to enable/disable prompt suggestions
434- Added `/settings` as an alias for the `/config` command
435- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
436- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
437- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
438- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
439- New syntax highlighting engine for native build
440
441## 2.0.70
442
443- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
444- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
445- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
446- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
447- Fixed input being cleared when processing queued commands while the user was typing
448- Fixed prompt suggestions replacing typed input when pressing Tab
449- Fixed diff view not updating when terminal is resized
450- Improved memory usage by 3x for large conversations
451- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
452- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
453- Fix thinking mode toggle in /config not persisting correctly
454- Improve UI for file creation permission dialog
455
456## 2.0.69
457
458- Minor bugfixes
459
460## 2.0.68
461
462- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
463- Fixed a bug where disallowed MCP tools were visible to the model
464- Fixed an issue where steering messages could be lost while a subagent is working
465- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
466- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
467- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
468
469## 2.0.67
470
471- Thinking mode is now enabled by default for Opus 4.5
472- Thinking mode configuration has moved to /config
473- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
474- Show reason why autoupdater is disabled in `/doctor`
475- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
476- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
477- Fixed scroll position resetting after deleting a permission rule in `/permissions`
478- 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
479- Fixed `claude install --force` not bypassing stale lock files
480- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
481- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
482
483## 2.0.65
484
485- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
486- Added context window information to status line input
487- Added `fileSuggestion` setting for custom `@` file search commands
488- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
489- Fixed prompt not being saved to history when aborting a query with Escape
490- Fixed Read tool image handling to identify format from bytes instead of file extension
491
492## 2.0.64
493
494- Made auto-compacting instant
495- Agents and bash commands can run asynchronously and send messages to wake up the main agent
496- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
497- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
498- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
499- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
500- Fixed auto-loading .env when using native installer
501- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
502- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
503- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
504- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
505- Bedrock: Improve efficiency of token counting
506- Bedrock: Add support for `aws login` AWS Management Console credentials
507- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
508
509## 2.0.62
510
511- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
512- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
513- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
514- Fixed slash command selection not working when multiple commands share the same name
515- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
516- Fixed running versions getting removed because lock file incorrectly going stale
517- Fixed IDE diff tab not closing when rejecting file changes
518
519## 2.0.61
520
521- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
522
523## 2.0.60
524
525- Added background agent support. Agents run in the background while you work
526- Added --disable-slash-commands CLI flag to disable all slash commands
527- Added model name to "Co-Authored-By" commit messages
528- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
529- Updated Fetch to skip summarization for pre-approved websites
530- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
531
532## 2.0.59
533
534- Added --agent CLI flag to override the agent setting for the current session
535- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
536- VS Code: Fixed .claude.json config file being read from incorrect location
537
538## 2.0.58
539
540- Pro users now have access to Opus 4.5 as part of their subscription!
541- Fixed timer duration showing "11m 60s" instead of "12m 0s"
542- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
543
544## 2.0.57
545
546- Added feedback input when rejecting plans, allowing users to tell Claude what to change
547- VSCode: Added streaming message support for real-time response display
548
549## 2.0.56
550
551- Added setting to enable/disable terminal progress bar (OSC 9;4)
552- 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.
553
554## 2.0.55
555
556- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
557- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
558- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
559- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
560
561## 2.0.54
562
563- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
564- Fix issue with excessive iTerm notifications
565
566## 2.0.52
567
568- Fixed duplicate message display when starting Claude with a command line argument
569- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
570- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
571- Permit some uses of `$!` in bash commands
572
573## 2.0.51
574
575- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
576- Introducing Claude Code for Desktop: https://claude.com/download
577- 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
578- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
579- Plan Mode now builds more precise plans and executes more thoroughly
580- Usage limit notifications now easier to understand
581- Switched `/usage` back to "% used"
582- Fixed handling of thinking errors
583- Fixed performance regression
584
585## 2.0.50
586
587- Fixed bug preventing calling MCP tools that have nested references in their input schemas
588- Silenced a noisy but harmless error during upgrades
589- Improved ultrathink text display
590- Improved clarity of 5-hour session limit warning message
591
592## 2.0.49
593
594- Added readline-style ctrl-y for pasting deleted text
595- Improved clarity of usage limit warning message
596- Fixed handling of subagent permissions
597
598## 2.0.47
599
600- Improved error messages and validation for `claude --teleport`
601- Improved error handling in `/usage`
602- Fixed race condition with history entry not getting logged at exit
603- Fixed Vertex AI configuration not being applied from `settings.json`
604
605## 2.0.46
606
607- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
608
609## 2.0.45
610
611- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
612- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
613- Send background tasks to Claude Code on the web by starting a message with `&`
614
615## 2.0.43
616
617- Added `permissionMode` field for custom agents
618- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
619- Added skills frontmatter field to declare skills to auto-load for subagents
620- Added the `SubagentStart` hook event
621- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
622- Fixed duplicate rendering of some messages in the UI
623- Fixed some visual flickers
624- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
625
626## 2.0.42
627
628- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
629
630## 2.0.41
631
632- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
633- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
634- Fixed incorrect labeling of user settings vs project settings in command descriptions
635- Fixed crash when plugin command hooks timeout during execution
636- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
637- Fixed broken security documentation links in trust dialogs and onboarding
638- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
639- ctrl-r history search landing on a slash command no longer cancels the search
640- SDK: Support custom timeouts for hooks
641- Allow more safe git commands to run without approval
642- Plugins: Added support for sharing and installing output styles
643- Teleporting a session from web will automatically set the upstream branch
644
645## 2.0.37
646
647- Fixed how idleness is computed for notifications
648- Hooks: Added matcher values for Notification hook events
649- Output Styles: Added `keep-coding-instructions` option to frontmatter
650
651## 2.0.36
652
653- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
654- Fixed queued messages being incorrectly executed as bash commands
655- Fixed input being lost when typing while a queued message is processed
656
657## 2.0.35
658
659- Improve fuzzy search results when searching commands
660- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
661- 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
662- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
663- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
664
665## 2.0.34
666
667- VSCode Extension: Added setting to configure the initial permission mode for new conversations
668- Improved file path suggestion performance with native Rust-based fuzzy finder
669- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
670- Fixed memory crash when reading or writing large files (especially base64-encoded images)
671
672## 2.0.33
673
674- Native binary installs now launch quicker.
675- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
676- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
677
678## 2.0.32
679
680- Un-deprecate output styles based on community feedback
681- Added `companyAnnouncements` setting for displaying announcements on startup
682- Fixed hook progress messages not updating correctly during PostToolUse hook execution
683
684## 2.0.31
685
686- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
687- Vertex: add support for Web Search on supported models
688- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
689- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
690- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
691- Fixed plugin uninstall not removing plugins
692
693## 2.0.30
694
695- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
696- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
697- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
698- Added prompt-based stop hooks
699- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
700- Enabled SSE MCP servers on native build
701- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
702- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
703- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
704- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
705- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
706- Fixed bug that saved session permissions to local settings
707- Fixed MCP tools not being available to sub-agents
708- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
709- Fixed delay when navigating through typeahead suggestions with arrow keys
710- VSCode: Restored selection indicator in input footer showing current file or code selection status
711
712## 2.0.28
713
714- Plan mode: introduced new Plan subagent
715- Subagents: claude can now choose to resume subagents
716- Subagents: claude can dynamically choose the model used by its subagents
717- SDK: added --max-budget-usd flag
718- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
719- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
720- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
721- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
722- Various other bug fixes
723
724## 2.0.27
725
726- New UI for permission prompts
727- Added current branch filtering and search to session resume screen for easier navigation
728- Fixed directory @-mention causing "No assistant message found" error
729- VSCode Extension: Add config setting to include .gitignored files in file searches
730- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
731
732## 2.0.25
733
734- 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
735
736## 2.0.24
737
738- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
739- Claude Code Web: Support for Web -> CLI teleport
740- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
741- Bedrock: Display awsAuthRefresh output when auth is required
742
743## 2.0.22
744
745- Fixed content layout shift when scrolling through slash commands
746- IDE: Add toggle to enable/disable thinking.
747- Fix bug causing duplicate permission prompts with parallel tool calls
748- Add support for enterprise managed MCP allowlist and denylist
749
750## 2.0.21
751
752- Support MCP `structuredContent` field in tool responses
753- Added an interactive question tool
754- Claude will now ask you questions more often in plan mode
755- Added Haiku 4.5 as a model option for Pro users
756- Fixed an issue where queued commands don't have access to previous messages' output
757
758## 2.0.20
759
760- Added support for Claude Skills
761
762## 2.0.19
763
764- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
765- Fixed a bug where Haiku was unnecessarily called in print mode
766
767## 2.0.17
768
769- Added Haiku 4.5 to model selector!
770- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
771- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
772- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
773- OTEL: support HTTP_PROXY and HTTPS_PROXY
774- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
775
776## 2.0.15
777
778- Fixed bug with resuming where previously created files needed to be read again before writing
779- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
780
781## 2.0.14
782
783- Fix @-mentioning MCP servers to toggle them on/off
784- Improve permission checks for bash with inline env vars
785- Fix ultrathink + thinking toggle
786- Reduce unnecessary logins
787- Document --system-prompt
788- Several improvements to rendering
789- Plugins UI polish
790
791## 2.0.13
792
793- Fixed `/plugin` not working on native build
794
795## 2.0.12
796
797- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
798- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
799- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
800- `/plugin validate` command for validating plugin structure and configuration
801- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
802- Plugin documentation available at https://code.claude.com/docs/en/plugins
803- Comprehensive error messages and diagnostics via `/doctor` command
804- Avoid flickering in `/model` selector
805- Improvements to `/help`
806- Avoid mentioning hooks in `/resume` summaries
807- Changes to the "verbose" setting in `/config` now persist across sessions
808
809## 2.0.11
810
811- Reduced system prompt size by 1.4k tokens
812- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
813- Fixed Opus fallback rate limit errors appearing incorrectly
814- Fixed /add-dir command selecting wrong default tab
815
816## 2.0.10
817
818- Rewrote terminal renderer for buttery smooth UI
819- Enable/disable MCP servers by @mentioning, or in /mcp
820- Added tab completion for shell commands in bash mode
821- PreToolUse hooks can now modify tool inputs
822- Press Ctrl-G to edit your prompt in your system's configured text editor
823- Fixes for bash permission checks with environment variables in the command
824
825## 2.0.9
826
827- Fix regression where bash backgrounding stopped working
828
829## 2.0.8
830
831- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
832- IDE: Add drag-and-drop support for files and folders in chat
833- /context: Fix counting for thinking blocks
834- Improve message rendering for users with light themes on dark terminals
835- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
836
837## 2.0.5
838
839- IDE: Fix IME unintended message submission with Enter and Tab
840- IDE: Add "Open in Terminal" link in login screen
841- Fix unhandled OAuth expiration 401 API errors
842- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
843
844## 2.0.1
845
846- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
847- Various bug fixes and presentation improvements
848
849## 2.0.0
850
851- New native VS Code extension
852- Fresh coat of paint throughout the whole app
853- /rewind a conversation to undo code changes
854- /usage command to see plan limits
855- Tab to toggle thinking (sticky across sessions)
856- Ctrl-R to search history
857- Unshipped claude config command
858- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
859- SDK: The Claude Code SDK is now the Claude Agent SDK
860- Add subagents dynamically with `--agents` flag
861
862## 1.0.126
863
864- Enable /context command for Bedrock and Vertex
865- Add mTLS support for HTTP-based OpenTelemetry exporters
866
867## 1.0.124
868
869- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
870- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
871- No longer inform Claude of the list of allowed tools when permission is denied
872- Fixed security vulnerability in Bash tool permission checks
873- Improved VSCode extension performance for large files
874
875## 1.0.123
876
877- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
878- Fixed thinking mode triggering on negation phrases like "don't think"
879- Fixed rendering performance degradation during token streaming
880- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
881- Enhanced BashTool environment snapshot logging
882- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
883- Migrated --debug logging to a file, to enable easy tailing & filtering
884
885## 1.0.120
886
887- Fix input lag during typing, especially noticeable with large prompts
888- Improved VSCode extension command registry and sessions dialog user experience
889- Enhanced sessions dialog responsiveness and visual feedback
890- Fixed IDE compatibility issue by removing worktree support check
891- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
892
893## 1.0.119
894
895- Fix Windows issue where process visually freezes on entering interactive mode
896- Support dynamic headers for MCP servers via headersHelper configuration
897- Fix thinking mode not working in headless sessions
898- Fix slash commands now properly update allowed tools instead of replacing them
899
900## 1.0.117
901
902- Add Ctrl-R history search to recall previous commands like bash/zsh
903- Fix input lag while typing, especially on Windows
904- Add sed command to auto-allowed commands in acceptEdits mode
905- Fix Windows PATH comparison to be case-insensitive for drive letters
906- Add permissions management hint to /add-dir output
907
908## 1.0.115
909
910- Improve thinking mode display with enhanced visual effects
911- Type /t to temporarily disable thinking mode in your prompt
912- Improve path validation for glob and grep tools
913- Show condensed output for post-tool hooks to reduce visual clutter
914- Fix visual feedback when loading state completes
915- Improve UI consistency for permission request dialogs
916
917## 1.0.113
918
919- Deprecated piped input in interactive mode
920- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
921
922## 1.0.112
923
924- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
925- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
926- Hooks: Added systemMessage support for SessionEnd hooks
927- Added `spinnerTipsEnabled` setting to disable spinner tips
928- IDE: Various improvements and bug fixes
929
930## 1.0.111
931
932- /model now validates provided model names
933- Fixed Bash tool crashes caused by malformed shell syntax parsing
934
935## 1.0.110
936
937- /terminal-setup command now supports WezTerm
938- MCP: OAuth tokens now proactively refresh before expiration
939- Fixed reliability issues with background Bash processes
940
941## 1.0.109
942
943- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
944
945## 1.0.106
946
947- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
948
949## 1.0.97
950
951- Settings: /doctor now validates permission rule syntax and suggests corrections
952
953## 1.0.94
954
955- Vertex: add support for global endpoints for supported models
956- /memory command now allows direct editing of all imported memory files
957- SDK: Add custom tools as callbacks
958- Added /todos command to list current todo items
959
960## 1.0.93
961
962- Windows: Add alt + v shortcut for pasting images from clipboard
963- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
964
965## 1.0.90
966
967- Settings file changes take effect immediately - no restart required
968
969## 1.0.88
970
971- Fixed issue causing "OAuth authentication is currently not supported"
972- Status line input now includes `exceeds_200k_tokens`
973- Fixed incorrect usage tracking in /cost.
974- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
975- Bedrock: Updated default Sonnet model to Sonnet 4
976
977## 1.0.86
978
979- Added /context to help users self-serve debug context issues
980- SDK: Added UUID support for all SDK messages
981- SDK: Added `--replay-user-messages` to replay user messages back to stdout
982
983## 1.0.85
984
985- Status line input now includes session cost info
986- Hooks: Introduced SessionEnd hook
987
988## 1.0.84
989
990- Fix tool_use/tool_result id mismatch error when network is unstable
991- Fix Claude sometimes ignoring real-time steering when wrapping up a task
992- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
993- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
994
995## 1.0.83
996
997- @-mention: Support files with spaces in path
998- New shimmering spinner
999
1000## 1.0.82
1001
1002- SDK: Add request cancellation support
1003- SDK: New additionalDirectories option to search custom paths, improved slash command processing
1004- Settings: Validation prevents invalid fields in .claude/settings.json files
1005- MCP: Improve tool name consistency
1006- Bash: Fix crash when Claude tries to automatically read large files
1007
1008## 1.0.81
1009
1010- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
1011- Agents: Fix custom agent loading when agent files are unparsable
1012
1013## 1.0.80
1014
1015- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
1016
1017## 1.0.77
1018
1019- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
1020- SDK: Add session support and permission denial tracking
1021- Fix token limit errors in conversation summarization
1022- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
1023
1024## 1.0.73
1025
1026- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
1027- MCP: Press Esc to cancel OAuth authentication flows
1028- Bash: Improved command validation and reduced false security warnings
1029- UI: Enhanced spinner animations and status line visual hierarchy
1030- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
1031
1032## 1.0.72
1033
1034- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
1035
1036## 1.0.71
1037
1038- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
1039- Customizable status line: add your terminal prompt to Claude Code with /statusline
1040
1041## 1.0.70
1042
1043- Performance: Optimized message rendering for better performance with large contexts
1044- Windows: Fixed native file search, ripgrep, and subagent functionality
1045- Added support for @-mentions in slash command arguments
1046
1047## 1.0.69
1048
1049- Upgraded Opus to version 4.1
1050
1051## 1.0.68
1052
1053- Fix incorrect model names being used for certain commands like `/pr-comments`
1054- 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.
1055- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
1056- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
1057- SDK: Added canUseTool callback support for tool confirmation
1058- Added `disableAllHooks` setting
1059- Improved file suggestions performance in large repos
1060
1061## 1.0.65
1062
1063- IDE: Fixed connection stability issues and error handling for diagnostics
1064- Windows: Fixed shell environment setup for users without .bashrc files
1065
1066## 1.0.64
1067
1068- Agents: Added model customization support - you can now specify which model an agent should use
1069- Agents: Fixed unintended access to the recursive agent tool
1070- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
1071- SDK: Fixed user input tracking across multi-turn conversations
1072- Added hidden files to file search and @-mention suggestions
1073
1074## 1.0.63
1075
1076- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
1077
1078## 1.0.62
1079
1080- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
1081- Hooks: Added SessionStart hook for new session initialization
1082- /add-dir command now supports typeahead for directory paths
1083- Improved network connectivity check reliability
1084
1085## 1.0.61
1086
1087- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
1088- Settings: Added `--settings` flag to load settings from a JSON file
1089- Settings: Fixed resolution of settings files paths that are symlinks
1090- OTEL: Fixed reporting of wrong organization after authentication changes
1091- Slash commands: Fixed permissions checking for allowed-tools with Bash
1092- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
1093- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
1094- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
1095
1096## 1.0.60
1097
1098- You can now create custom subagents for specialized tasks! Run /agents to get started
1099
1100## 1.0.59
1101
1102- SDK: Added tool confirmation support with canUseTool callback
1103- SDK: Allow specifying env for spawned process
1104- Hooks: Exposed PermissionDecision to hooks (including "ask")
1105- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
1106- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
1107
1108## 1.0.58
1109
1110- Added support for reading PDFs
1111- MCP: Improved server health status display in 'claude mcp list'
1112- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
1113
1114## 1.0.57
1115
1116- Added support for specifying a model in slash commands
1117- Improved permission messages to help Claude understand allowed tools
1118- Fix: Remove trailing newlines from bash output in terminal wrapping
1119
1120## 1.0.56
1121
1122- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
1123- Fixes for WSL IDE detection
1124- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
1125
1126## 1.0.55
1127
1128- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
1129- Windows: fixed Ctrl+Z crash
1130- SDK: Added ability to capture error logging
1131- Add --system-prompt-file option to override system prompt in print mode
1132
1133## 1.0.54
1134
1135- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
1136- Custom slash commands: Added argument-hint to frontmatter
1137- Windows: OAuth uses port 45454 and properly constructs browser URL
1138- Windows: mode switching now uses alt + m, and plan mode renders properly
1139- Shell: Switch to in-memory shell snapshot to fix file-related errors
1140
1141## 1.0.53
1142
1143- Updated @-mention file truncation from 100 lines to 2000 lines
1144- 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).
1145
1146## 1.0.52
1147
1148- Added support for MCP server instructions
1149
1150## 1.0.51
1151
1152- Added support for native Windows (requires Git for Windows)
1153- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
1154- Settings: /doctor can now help you identify and fix invalid setting files
1155- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
1156- Increased auto-compact warning threshold from 60% to 80%
1157- Fixed an issue with handling user directories with spaces for shell snapshots
1158- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
1159- Custom slash commands: Fixed user-level commands in subdirectories
1160- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
1161
1162## 1.0.48
1163
1164- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
1165- Added progress messages to Bash tool based on the last 5 lines of command output
1166- Added expanding variables support for MCP server configuration
1167- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
1168- Improved IDE extension path handling when Claude Code runs in WSL
1169- Hooks: Added a PreCompact hook
1170- Vim mode: Added c, f/F, t/T
1171
1172## 1.0.45
1173
1174- Redesigned Search (Grep) tool with new tool input parameters and features
1175- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
1176- Fixed config file corruption issue by enforcing atomic writes
1177- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
1178- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
1179- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
1180
1181## 1.0.44
1182
1183- New /export command lets you quickly export a conversation for sharing
1184- MCP: resource_link tool results are now supported
1185- MCP: tool annotations and tool titles now display in /mcp view
1186- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
1187
1188## 1.0.43
1189
1190- Fixed a bug where the theme selector was saving excessively
1191- Hooks: Added EPIPE system error handling
1192
1193## 1.0.42
1194
1195- Added tilde (`~`) expansion support to `/add-dir` command
1196
1197## 1.0.41
1198
1199- Hooks: Split Stop hook triggering into Stop and SubagentStop
1200- Hooks: Enabled optional timeout configuration for each command
1201- Hooks: Added "hook_event_name" to hook input
1202- Fixed a bug where MCP tools would display twice in tool list
1203- New tool parameters JSON for Bash tool in `tool_decision` event
1204
1205## 1.0.40
1206
1207- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1208
1209## 1.0.39
1210
1211- New Active Time metric in OpenTelemetry logging
1212
1213## 1.0.38
1214
1215- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1216
1217## 1.0.37
1218
1219- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1220
1221## 1.0.36
1222
1223- Web search now takes today's date into context
1224- Fixed a bug where stdio MCP servers were not terminating properly on exit
1225
1226## 1.0.35
1227
1228- Added support for MCP OAuth Authorization Server discovery
1229
1230## 1.0.34
1231
1232- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1233
1234## 1.0.33
1235
1236- Improved logging functionality with session ID support
1237- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1238- Improvements to plan mode
1239
1240## 1.0.32
1241
1242- Updated loopback config for litellm
1243- Added forceLoginMethod setting to bypass login selection screen
1244
1245## 1.0.31
1246
1247- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1248
1249## 1.0.30
1250
1251- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1252- Improved file path autocomplete with filename matching
1253- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1254- Enhanced jq regex support for complex filters with pipes and select
1255
1256## 1.0.29
1257
1258- Improved CJK character support in cursor navigation and rendering
1259
1260## 1.0.28
1261
1262- Slash commands: Fix selector display during history navigation
1263- Resizes images before upload to prevent API size limit errors
1264- Added XDG_CONFIG_HOME support to configuration directory
1265- Performance optimizations for memory usage
1266- New attributes (terminal.type, language) in OpenTelemetry logging
1267
1268## 1.0.27
1269
1270- Streamable HTTP MCP servers are now supported
1271- Remote MCP servers (SSE and HTTP) now support OAuth
1272- MCP resources can now be @-mentioned
1273- /resume slash command to switch conversations within Claude Code
1274
1275## 1.0.25
1276
1277- Slash commands: moved "project" and "user" prefixes to descriptions
1278- Slash commands: improved reliability for command discovery
1279- Improved support for Ghostty
1280- Improved web search reliability
1281
1282## 1.0.24
1283
1284- Improved /mcp output
1285- Fixed a bug where settings arrays got overwritten instead of merged
1286
1287## 1.0.23
1288
1289- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1290- Released Python SDK: pip install claude-code-sdk to get started
1291
1292## 1.0.22
1293
1294- SDK: Renamed `total_cost` to `total_cost_usd`
1295
1296## 1.0.21
1297
1298- Improved editing of files with tab-based indentation
1299- Fix for tool_use without matching tool_result errors
1300- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1301
1302## 1.0.18
1303
1304- Added --add-dir CLI argument for specifying additional working directories
1305- Added streaming input support without require -p flag
1306- Improved startup performance and session storage performance
1307- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1308- Added detailed MCP server tools display (/mcp)
1309- MCP authentication and permission improvements
1310- Added auto-reconnection for MCP SSE connections on disconnect
1311- Fixed issue where pasted content was lost when dialogs appeared
1312
1313## 1.0.17
1314
1315- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1316- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1317- MCP server list UI improvements
1318- Update Claude Code process title to display "claude" instead of "node"
1319
1320## 1.0.11
1321
1322- Claude Code can now also be used with a Claude Pro subscription
1323- Added /upgrade for smoother switching to Claude Max plans
1324- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1325- Improved shell configuration error handling
1326- Improved todo list handling during compaction
1327
1328## 1.0.10
1329
1330- Added markdown table support
1331- Improved streaming performance
1332
1333## 1.0.8
1334
1335- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1336- Increased default otel interval from 1s -> 5s
1337- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1338- Fixed a regression where search tools unnecessarily asked for permissions
1339- Added support for triggering thinking non-English languages
1340- Improved compacting UI
1341
1342## 1.0.7
1343
1344- Renamed /allowed-tools -> /permissions
1345- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1346- Deprecated claude config commands in favor of editing settings.json
1347- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1348- Improved error handling for /install-github-app
1349- Bugfixes, UI polish, and tool reliability improvements
1350
1351## 1.0.6
1352
1353- Improved edit reliability for tab-indented files
1354- Respect CLAUDE_CONFIG_DIR everywhere
1355- Reduced unnecessary tool permission prompts
1356- Added support for symlinks in @file typeahead
1357- Bugfixes, UI polish, and tool reliability improvements
1358
1359## 1.0.4
1360
1361- Fixed a bug where MCP tool errors weren't being parsed correctly
1362
1363## 1.0.1
1364
1365- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1366- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1367- Updated documentation links and OAuth process descriptions
1368
1369## 1.0.0
1370
1371- Claude Code is now generally available
1372- Introducing Sonnet 4 and Opus 4 models
1373
1374## 0.2.125
1375
1376- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1377- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1378
1379## 0.2.117
1380
1381- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1382- Introduced settings.cleanupPeriodDays
1383- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1384- Introduced --debug mode
1385
1386## 0.2.108
1387
1388- You can now send messages to Claude while it works to steer Claude in real-time
1389- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1390- Fixed a bug where thinking was not working in -p mode
1391- Fixed a regression in /cost reporting
1392- Deprecated MCP wizard interface in favor of other MCP commands
1393- Lots of other bugfixes and improvements
1394
1395## 0.2.107
1396
1397- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1398
1399## 0.2.106
1400
1401- MCP SSE server configs can now specify custom headers
1402- Fixed a bug where MCP permission prompt didn't always show correctly
1403
1404## 0.2.105
1405
1406- Claude can now search the web
1407- Moved system & account status to /status
1408- Added word movement keybindings for Vim
1409- Improved latency for startup, todo tool, and file edits
1410
1411## 0.2.102
1412
1413- Improved thinking triggering reliability
1414- Improved @mention reliability for images and folders
1415- You can now paste multiple large chunks into one prompt
1416
1417## 0.2.100
1418
1419- Fixed a crash caused by a stack overflow error
1420- Made db storage optional; missing db support disables --continue and --resume
1421
1422## 0.2.98
1423
1424- Fixed an issue where auto-compact was running twice
1425
1426## 0.2.96
1427
1428- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1429
1430## 0.2.93
1431
1432- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1433- Claude now has access to a Todo list that helps it stay on track and be more organized
1434
1435## 0.2.82
1436
1437- Added support for --disallowedTools
1438- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1439
1440## 0.2.75
1441
1442- Hit Enter to queue up additional messages while Claude is working
1443- Drag in or copy/paste image files directly into the prompt
1444- @-mention files to directly add them to context
1445- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1446- Improved performance for filename auto-complete
1447
1448## 0.2.74
1449
1450- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1451- Task tool can now perform writes and run bash commands
1452
1453## 0.2.72
1454
1455- Updated spinner to indicate tokens loaded and tool usage
1456
1457## 0.2.70
1458
1459- Network commands like curl are now available for Claude to use
1460- Claude can now run multiple web queries in parallel
1461- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1462
1463## 0.2.69
1464
1465- Fixed UI glitches with improved Select component behavior
1466- Enhanced terminal output display with better text truncation logic
1467
1468## 0.2.67
1469
1470- Shared project permission rules can be saved in .claude/settings.json
1471
1472## 0.2.66
1473
1474- Print mode (-p) now supports streaming output via --output-format=stream-json
1475- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1476
1477## 0.2.63
1478
1479- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1480
1481## 0.2.61
1482
1483- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1484- Enhanced image detection for more reliable clipboard paste functionality
1485- Fixed an issue where ESC key could crash the conversation history selector
1486
1487## 0.2.59
1488
1489- Copy+paste images directly into your prompt
1490- Improved progress indicators for bash and fetch tools
1491- Bugfixes for non-interactive mode (-p)
1492
1493## 0.2.54
1494
1495- Quickly add to Memory by starting your message with '#'
1496- Press ctrl+r to see full output for long tool results
1497- Added support for MCP SSE transport
1498
1499## 0.2.53
1500
1501- New web fetch tool lets Claude view URLs that you paste in
1502- Fixed a bug with JPEG detection
1503
1504## 0.2.50
1505
1506- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1507
1508## 0.2.49
1509
1510- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1511
1512## 0.2.47
1513
1514- Press Tab to auto-complete file and folder names
1515- Press Shift + Tab to toggle auto-accept for file edits
1516- Automatic conversation compaction for infinite conversation length (toggle with /config)
1517
1518## 0.2.44
1519
1520- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1521
1522## 0.2.41
1523
1524- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1525- MCP server startup no longer blocks the app from starting up
1526
1527## 0.2.37
1528
1529- New /release-notes command lets you view release notes at any time
1530- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1531
1532## 0.2.36
1533
1534- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1535- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1536
1537## 0.2.34
1538
1539- Vim bindings for text input - enable with /vim or /config
1540
1541## 0.2.32
1542
1543- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1544- Fix for some PersistentShell issues
1545
1546## 0.2.31
1547
1548- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1549- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1550
1551## 0.2.30
1552
1553- Added ANSI color theme for better terminal compatibility
1554- Fixed issue where slash command arguments weren't being sent properly
1555- (Mac-only) API keys are now stored in macOS Keychain
1556
1557## 0.2.26
1558
1559- New /approved-tools command for managing tool permissions
1560- Word-level diff display for improved code readability
1561- Fuzzy matching for slash commands
1562
1563## 0.2.21
1564
1565- Fuzzy matching for /commands