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
1,486 lines · 74.7 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.19
4
5- Added env var `CLAUDE_CODE_ENABLE_TASKS`, set to `false` to keep the old system temporarily
6- Added shorthand `$0`, `$1`, etc. for accessing individual arguments in custom commands
7- Fixed crashes on processors without AVX instruction support
8- Fixed subscription intent not being preserved through OAuth authentication flow, causing users who selected the Pro plan to land on the regular interface instead of completing the upgrade after sign-in
9- Fixed dangling Claude Code processes when terminal is closed by catching EIO errors from `process.exit()` and using SIGKILL as fallback
10- Fixed `/rename` and `/tag` not updating the correct session when resuming from a different directory (e.g., git worktrees)
11- Fixed resuming sessions by custom title not working when run from a different directory
12- Fixed pasted text content being lost when using prompt stash (Ctrl+S) and restore
13- Fixed agent list displaying "Sonnet (default)" instead of "Inherit (default)" for agents without an explicit model setting
14- Fixed backgrounded hook commands not returning early, potentially causing the session to wait on a process that was intentionally backgrounded
15- Fixed file write preview omitting empty lines
16- Changed skills without additional permissions or hooks to be allowed without requiring approval
17- Changed indexed argument syntax from `$ARGUMENTS.0` to `$ARGUMENTS[0]` (bracket syntax)
18- [SDK] Added replay of `queued_command` attachment messages as `SDKUserMessageReplay` events when `replayUserMessages` is enabled
19- [VSCode] Enabled session forking and rewind functionality for all users
20
21## 2.1.18
22
23- 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
24
25## 2.1.17
26
27- Fixed crashes on processors without AVX instruction support
28
29## 2.1.16
30
31- Added new task management system, including new capabilities like dependency tracking
32- [VSCode] Added native plugin management support
33- [VSCode] Added ability for OAuth users to browse and resume remote Claude sessions from the Sessions dialog
34- Fixed out-of-memory crashes when resuming sessions with heavy subagent usage
35- Fixed an issue where the "context remaining" warning was not hidden after running `/compact`
36- Fixed session titles on the resume screen not respecting the user's language setting
37- [IDE] Fixed a race condition on Windows where the Claude Code sidebar view container would not appear on start
38
39## 2.1.15
40
41- Added deprecation notification for npm installations - run `claude install` or see https://docs.anthropic.com/en/docs/claude-code/getting-started for more options
42- Improved UI rendering performance with React Compiler
43- Fixed the "Context left until auto-compact" warning not disappearing after running `/compact`
44- Fixed MCP stdio server timeout not killing child process, which could cause UI freezes
45
46## 2.1.14
47
48- Added history-based autocomplete in bash mode (`!`) - type a partial command and press Tab to complete from your bash command history
49- Added search to installed plugins list - type to filter by name or description
50- Added support for pinning plugins to specific git commit SHAs, allowing marketplace entries to install exact versions
51- Fixed a regression where the context window blocking limit was calculated too aggressively, blocking users at ~65% context usage instead of the intended ~98%
52- Fixed memory issues that could cause crashes when running parallel subagents
53- Fixed memory leak in long-running sessions where stream resources were not cleaned up after shell commands completed
54- Fixed `@` symbol incorrectly triggering file autocomplete suggestions in bash mode
55- Fixed `@`-mention menu folder click behavior to navigate into directories instead of selecting them
56- Fixed `/feedback` command generating invalid GitHub issue URLs when description is very long
57- Fixed `/context` command to show the same token count and percentage as the status line in verbose mode
58- Fixed an issue where `/config`, `/context`, `/model`, and `/todos` command overlays could close unexpectedly
59- Fixed slash command autocomplete selecting wrong command when typing similar commands (e.g., `/context` vs `/compact`)
60- Fixed inconsistent back navigation in plugin marketplace when only one marketplace is configured
61- Fixed iTerm2 progress bar not clearing properly on exit, preventing lingering indicators and bell sounds
62- Improved backspace to delete pasted text as a single token instead of one character at a time
63- [VSCode] Added `/usage` command to display current plan usage
64
65## 2.1.12
66
67- Fixed message rendering bug
68
69## 2.1.11
70
71- Fixed excessive MCP connection requests for HTTP/SSE transports
72
73## 2.1.10
74
75- Added new `Setup` hook event that can be triggered via `--init`, `--init-only`, or `--maintenance` CLI flags for repository setup and maintenance operations
76- Added keyboard shortcut 'c' to copy OAuth URL when browser doesn't open automatically during login
77- Fixed a crash when running bash commands containing heredocs with JavaScript template literals like `${index + 1}`
78- Improved startup to capture keystrokes typed before the REPL is fully ready
79- Improved file suggestions to show as removable attachments instead of inserting text when accepted
80- [VSCode] Added install count display to plugin listings
81- [VSCode] Added trust warning when installing plugins
82
83## 2.1.9
84
85- Added `auto:N` syntax for configuring the MCP tool search auto-enable threshold, where N is the context window percentage (0-100)
86- Added `plansDirectory` setting to customize where plan files are stored
87- Added external editor support (Ctrl+G) in AskUserQuestion "Other" input field
88- Added session URL attribution to commits and PRs created from web sessions
89- Added support for `PreToolUse` hooks to return `additionalContext` to the model
90- Added `${CLAUDE_SESSION_ID}` string substitution for skills to access the current session ID
91- Fixed long sessions with parallel tool calls failing with an API error about orphan tool_result blocks
92- Fixed MCP server reconnection hanging when cached connection promise never resolves
93- Fixed Ctrl+Z suspend not working in terminals using Kitty keyboard protocol (Ghostty, iTerm2, kitty, WezTerm)
94
95## 2.1.7
96
97- Added `showTurnDuration` setting to hide turn duration messages (e.g., "Cooked for 1m 6s")
98- Added ability to provide feedback when accepting permission prompts
99- Added inline display of agent's final response in task notifications, making it easier to see results without reading the full transcript file
100- Fixed security vulnerability where wildcard permission rules could match compound commands containing shell operators
101- Fixed false "file modified" errors on Windows when cloud sync tools, antivirus scanners, or Git touch file timestamps without changing content
102- Fixed orphaned tool_result errors when sibling tools fail during streaming execution
103- 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)
104- Fixed spinner briefly flashing when running local slash commands like `/model` or `/theme`
105- Fixed terminal title animation jitter by using fixed-width braille characters
106- Fixed plugins with git submodules not being fully initialized when installed
107- Fixed bash commands failing on Windows when temp directory paths contained characters like `t` or `n` that were misinterpreted as escape sequences
108- Improved typing responsiveness by reducing memory allocation overhead in terminal rendering
109- 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.
110- Changed OAuth and API Console URLs from console.anthropic.com to platform.claude.com
111- [VSCode] Fixed `claudeProcessWrapper` setting passing the wrapper path instead of the Claude binary path
112
113## 2.1.6
114
115- Added search functionality to `/config` command for quickly filtering settings
116- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
117- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
118- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
119- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
120- Added an error display when the editor fails during Ctrl+G
121- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
122- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
123- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
124- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
125- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
126- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
127- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
128- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
129- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
130- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
131- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
132- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
133- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
134- Improved @ autocomplete with icons for different suggestion types and single-line formatting
135- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
136- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
137- Changed terminal title to "Claude Code" on startup for better window identification
138- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
139- [VSCode] Fixed usage indicator not updating after manual compact
140
141## 2.1.5
142
143- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
144
145## 2.1.4
146
147- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
148- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
149
150## 2.1.3
151
152- Merged slash commands and skills, simplifying the mental model with no change in behavior
153- Added release channel (`stable` or `latest`) toggle to `/config`
154- 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
155- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
156- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
157- Fixed mismatch between background task count in status bar and items shown in tasks dialog
158- Fixed sub-agents using the wrong model during conversation compaction
159- Fixed web search in sub-agents using incorrect model
160- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
161- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
162- Improved slash command suggestion readability by truncating long descriptions to 2 lines
163- Changed tool hook execution timeout from 60 seconds to 10 minutes
164- [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)
165
166## 2.1.2
167
168- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
169- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
170- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
171- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
172- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
173- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
174- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
175- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
176- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
177- Fixed updates incorrectly claiming another installation is in progress
178- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
179- Fixed remote session URL and teleport being broken when using `/tasks` command
180- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
181- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
182- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
183- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
184- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
185- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
186- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
187- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
188- [SDK] Changed minimum zod peer dependency to ^4.0.0
189- [VSCode] Fixed usage display not updating after manual compact
190
191## 2.1.0
192
193- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
194- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
195- Added support for `agent` field in skills to specify agent type for execution
196- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
197- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
198- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
199- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
200- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
201- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
202- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
203- Fixed Esc key with queued prompts to only move them to input without canceling the running task
204- Reduced permission prompts for complex bash commands
205- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
206- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
207- Fixed plugin path resolution for file-based marketplace sources
208- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
209- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
210- Fixed Claude in Chrome support for WSL environments
211- Fixed Windows native installer silently failing when executable creation fails
212- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
213- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
214- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
215- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
216- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
217- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
218- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
219- Added hooks support for skill and slash command frontmatter
220- 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
221- Added `/plan` command shortcut to enable plan mode directly from the prompt
222- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
223- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
224- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
225- Added support for `once: true` config for hooks
226- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
227- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
228- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
229- Added left/right arrow key navigation for cycling through tabs in dialogs
230- Added real-time thinking block display in Ctrl+O transcript mode
231- Added filepath to full output in background bash task details dialog
232- Added Skills as a separate category in the context visualization
233- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
234- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
235- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
236- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
237- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
238- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
239- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
240- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
241- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
242- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
243- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
244- Fixed commands with `$()` command substitution failing with parse errors
245- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
246- 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)
247- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
248- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
249- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
250- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
251- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
252- Fixed subagents sometimes not inheriting the parent's model by default
253- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
254- Fixed duplicate Bash commands appearing in permission request option labels
255- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
256- Fixed background task completion notifications to appear proactively with bullet point
257- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
258- Fixed cursor disappearing after dismissing permission dialogs
259- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
260- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
261- Fixed images being silently dropped when queueing messages while backgrounding a task
262- Fixed large pasted images failing with "Image was too large" error
263- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
264- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
265- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
266- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
267- Fixed spinner token counter to properly accumulate tokens from subagents during execution
268- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
269- Fixed race condition where LSP tool could return "no server available" during startup
270- Fixed feedback submission hanging indefinitely when network requests timeout
271- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
272- Fixed hook success message showing trailing colon when hook has no output
273- Multiple optimizations to improve startup performance
274- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
275- Improved performance when reading Jupyter notebooks with many cells
276- Improved reliability for piped input like `cat refactor.md | claude`
277- Improved reliability for AskQuestion tool
278- Improved sed in-place edit commands to render as file edits with diff preview
279- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
280- Improved compaction reliability
281- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
282- Improved skills to show progress while executing, displaying tool uses as they happen
283- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
284- Improved skill suggestions to prioritize recently and frequently used skills
285- Improved spinner feedback when waiting for the first response token
286- Improved token count display in spinner to include tokens from background agents
287- Improved incremental output for async agents to give the main thread more control and visibility
288- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
289- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
290- Improved macOS screenshot paste reliability with TIFF format support
291- Improved `/stats` output
292- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
293- Changed "Interrupted" message color from red to grey for a less alarming appearance
294- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
295- Removed underline styling from image reference links
296- [SDK] Changed minimum zod peer dependency to ^4.0.0
297- [VSCode] Added currently selected model name to the context menu
298- [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")
299- [VSCode] Fixed paragraph breaks not rendering in markdown content
300- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
301- [Windows] Fixed issue with improper rendering
302
303## 2.0.76
304
305- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
306
307## 2.0.75
308
309- Minor bugfixes
310
311## 2.0.74
312
313- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
314- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
315- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
316- Added syntax highlighting info to theme picker
317- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
318- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
319- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
320- Fixed a potential crash when syntax highlighting isn't initialized correctly
321- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
322- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
323- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
324- [Windows] Fixed issue with improper rendering
325- [VSCode] Added gift tag pictogram for year-end promotion message
326
327## 2.0.73
328
329- Added clickable `[Image #N]` links that open attached images in the default viewer
330- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
331- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
332- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
333- Fixed slow input history cycling and race condition that could overwrite text after message submission
334- Improved `/theme` command to open theme picker directly
335- Improved theme picker UI
336- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
337- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
338
339## 2.0.72
340
341- 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
342- Reduced terminal flickering
343- Added scannable QR code to mobile app tip for quick app downloads
344- Added loading indicator when resuming conversations for better feedback
345- Fixed `/context` command not respecting custom system prompts in non-interactive mode
346- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
347- Improved @ mention file suggestion speed (~3x faster in git repositories)
348- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
349- Improved settings validation errors to be more prominent
350- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
351
352## 2.0.71
353
354- Added /config toggle to enable/disable prompt suggestions
355- Added `/settings` as an alias for the `/config` command
356- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
357- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
358- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
359- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
360- New syntax highlighting engine for native build
361
362## 2.0.70
363
364- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
365- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
366- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
367- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
368- Fixed input being cleared when processing queued commands while the user was typing
369- Fixed prompt suggestions replacing typed input when pressing Tab
370- Fixed diff view not updating when terminal is resized
371- Improved memory usage by 3x for large conversations
372- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
373- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
374- Fix thinking mode toggle in /config not persisting correctly
375- Improve UI for file creation permission dialog
376
377## 2.0.69
378
379- Minor bugfixes
380
381## 2.0.68
382
383- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
384- Fixed a bug where disallowed MCP tools were visible to the model
385- Fixed an issue where steering messages could be lost while a subagent is working
386- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
387- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
388- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
389
390## 2.0.67
391
392- Thinking mode is now enabled by default for Opus 4.5
393- Thinking mode configuration has moved to /config
394- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
395- Show reason why autoupdater is disabled in `/doctor`
396- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
397- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
398- Fixed scroll position resetting after deleting a permission rule in `/permissions`
399- 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
400- Fixed `claude install --force` not bypassing stale lock files
401- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
402- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
403
404## 2.0.65
405
406- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
407- Added context window information to status line input
408- Added `fileSuggestion` setting for custom `@` file search commands
409- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
410- Fixed prompt not being saved to history when aborting a query with Escape
411- Fixed Read tool image handling to identify format from bytes instead of file extension
412
413## 2.0.64
414
415- Made auto-compacting instant
416- Agents and bash commands can run asynchronously and send messages to wake up the main agent
417- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
418- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
419- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
420- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
421- Fixed auto-loading .env when using native installer
422- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
423- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
424- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
425- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
426- Bedrock: Improve efficiency of token counting
427- Bedrock: Add support for `aws login` AWS Management Console credentials
428- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
429
430## 2.0.62
431
432- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
433- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
434- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
435- Fixed slash command selection not working when multiple commands share the same name
436- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
437- Fixed running versions getting removed because lock file incorrectly going stale
438- Fixed IDE diff tab not closing when rejecting file changes
439
440## 2.0.61
441
442- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
443
444## 2.0.60
445
446- Added background agent support. Agents run in the background while you work
447- Added --disable-slash-commands CLI flag to disable all slash commands
448- Added model name to "Co-Authored-By" commit messages
449- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
450- Updated Fetch to skip summarization for pre-approved websites
451- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
452
453## 2.0.59
454
455- Added --agent CLI flag to override the agent setting for the current session
456- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
457- VS Code: Fixed .claude.json config file being read from incorrect location
458
459## 2.0.58
460
461- Pro users now have access to Opus 4.5 as part of their subscription!
462- Fixed timer duration showing "11m 60s" instead of "12m 0s"
463- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
464
465## 2.0.57
466
467- Added feedback input when rejecting plans, allowing users to tell Claude what to change
468- VSCode: Added streaming message support for real-time response display
469
470## 2.0.56
471
472- Added setting to enable/disable terminal progress bar (OSC 9;4)
473- 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.
474
475## 2.0.55
476
477- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
478- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
479- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
480- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
481
482## 2.0.54
483
484- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
485- Fix issue with excessive iTerm notifications
486
487## 2.0.52
488
489- Fixed duplicate message display when starting Claude with a command line argument
490- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
491- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
492- Permit some uses of `$!` in bash commands
493
494## 2.0.51
495
496- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
497- Introducing Claude Code for Desktop: https://claude.com/download
498- 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
499- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
500- Plan Mode now builds more precise plans and executes more thoroughly
501- Usage limit notifications now easier to understand
502- Switched `/usage` back to "% used"
503- Fixed handling of thinking errors
504- Fixed performance regression
505
506## 2.0.50
507
508- Fixed bug preventing calling MCP tools that have nested references in their input schemas
509- Silenced a noisy but harmless error during upgrades
510- Improved ultrathink text display
511- Improved clarity of 5-hour session limit warning message
512
513## 2.0.49
514
515- Added readline-style ctrl-y for pasting deleted text
516- Improved clarity of usage limit warning message
517- Fixed handling of subagent permissions
518
519## 2.0.47
520
521- Improved error messages and validation for `claude --teleport`
522- Improved error handling in `/usage`
523- Fixed race condition with history entry not getting logged at exit
524- Fixed Vertex AI configuration not being applied from `settings.json`
525
526## 2.0.46
527
528- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
529
530## 2.0.45
531
532- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
533- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
534- Send background tasks to Claude Code on the web by starting a message with `&`
535
536## 2.0.43
537
538- Added `permissionMode` field for custom agents
539- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
540- Added skills frontmatter field to declare skills to auto-load for subagents
541- Added the `SubagentStart` hook event
542- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
543- Fixed duplicate rendering of some messages in the UI
544- Fixed some visual flickers
545- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
546
547## 2.0.42
548
549- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
550
551## 2.0.41
552
553- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
554- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
555- Fixed incorrect labeling of user settings vs project settings in command descriptions
556- Fixed crash when plugin command hooks timeout during execution
557- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
558- Fixed broken security documentation links in trust dialogs and onboarding
559- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
560- ctrl-r history search landing on a slash command no longer cancels the search
561- SDK: Support custom timeouts for hooks
562- Allow more safe git commands to run without approval
563- Plugins: Added support for sharing and installing output styles
564- Teleporting a session from web will automatically set the upstream branch
565
566## 2.0.37
567
568- Fixed how idleness is computed for notifications
569- Hooks: Added matcher values for Notification hook events
570- Output Styles: Added `keep-coding-instructions` option to frontmatter
571
572## 2.0.36
573
574- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
575- Fixed queued messages being incorrectly executed as bash commands
576- Fixed input being lost when typing while a queued message is processed
577
578## 2.0.35
579
580- Improve fuzzy search results when searching commands
581- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
582- 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
583- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
584- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
585
586## 2.0.34
587
588- VSCode Extension: Added setting to configure the initial permission mode for new conversations
589- Improved file path suggestion performance with native Rust-based fuzzy finder
590- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
591- Fixed memory crash when reading or writing large files (especially base64-encoded images)
592
593## 2.0.33
594
595- Native binary installs now launch quicker.
596- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
597- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
598
599## 2.0.32
600
601- Un-deprecate output styles based on community feedback
602- Added `companyAnnouncements` setting for displaying announcements on startup
603- Fixed hook progress messages not updating correctly during PostToolUse hook execution
604
605## 2.0.31
606
607- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
608- Vertex: add support for Web Search on supported models
609- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
610- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
611- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
612- Fixed plugin uninstall not removing plugins
613
614## 2.0.30
615
616- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
617- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
618- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
619- Added prompt-based stop hooks
620- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
621- Enabled SSE MCP servers on native build
622- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
623- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
624- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
625- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
626- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
627- Fixed bug that saved session permissions to local settings
628- Fixed MCP tools not being available to sub-agents
629- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
630- Fixed delay when navigating through typeahead suggestions with arrow keys
631- VSCode: Restored selection indicator in input footer showing current file or code selection status
632
633## 2.0.28
634
635- Plan mode: introduced new Plan subagent
636- Subagents: claude can now choose to resume subagents
637- Subagents: claude can dynamically choose the model used by its subagents
638- SDK: added --max-budget-usd flag
639- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
640- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
641- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
642- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
643- Various other bug fixes
644
645## 2.0.27
646
647- New UI for permission prompts
648- Added current branch filtering and search to session resume screen for easier navigation
649- Fixed directory @-mention causing "No assistant message found" error
650- VSCode Extension: Add config setting to include .gitignored files in file searches
651- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
652
653## 2.0.25
654
655- 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
656
657## 2.0.24
658
659- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
660- Claude Code Web: Support for Web -> CLI teleport
661- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
662- Bedrock: Display awsAuthRefresh output when auth is required
663
664## 2.0.22
665
666- Fixed content layout shift when scrolling through slash commands
667- IDE: Add toggle to enable/disable thinking.
668- Fix bug causing duplicate permission prompts with parallel tool calls
669- Add support for enterprise managed MCP allowlist and denylist
670
671## 2.0.21
672
673- Support MCP `structuredContent` field in tool responses
674- Added an interactive question tool
675- Claude will now ask you questions more often in plan mode
676- Added Haiku 4.5 as a model option for Pro users
677- Fixed an issue where queued commands don't have access to previous messages' output
678
679## 2.0.20
680
681- Added support for Claude Skills
682
683## 2.0.19
684
685- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
686- Fixed a bug where Haiku was unnecessarily called in print mode
687
688## 2.0.17
689
690- Added Haiku 4.5 to model selector!
691- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
692- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
693- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
694- OTEL: support HTTP_PROXY and HTTPS_PROXY
695- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
696
697## 2.0.15
698
699- Fixed bug with resuming where previously created files needed to be read again before writing
700- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
701
702## 2.0.14
703
704- Fix @-mentioning MCP servers to toggle them on/off
705- Improve permission checks for bash with inline env vars
706- Fix ultrathink + thinking toggle
707- Reduce unnecessary logins
708- Document --system-prompt
709- Several improvements to rendering
710- Plugins UI polish
711
712## 2.0.13
713
714- Fixed `/plugin` not working on native build
715
716## 2.0.12
717
718- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
719- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
720- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
721- `/plugin validate` command for validating plugin structure and configuration
722- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
723- Plugin documentation available at https://code.claude.com/docs/en/plugins
724- Comprehensive error messages and diagnostics via `/doctor` command
725- Avoid flickering in `/model` selector
726- Improvements to `/help`
727- Avoid mentioning hooks in `/resume` summaries
728- Changes to the "verbose" setting in `/config` now persist across sessions
729
730## 2.0.11
731
732- Reduced system prompt size by 1.4k tokens
733- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
734- Fixed Opus fallback rate limit errors appearing incorrectly
735- Fixed /add-dir command selecting wrong default tab
736
737## 2.0.10
738
739- Rewrote terminal renderer for buttery smooth UI
740- Enable/disable MCP servers by @mentioning, or in /mcp
741- Added tab completion for shell commands in bash mode
742- PreToolUse hooks can now modify tool inputs
743- Press Ctrl-G to edit your prompt in your system's configured text editor
744- Fixes for bash permission checks with environment variables in the command
745
746## 2.0.9
747
748- Fix regression where bash backgrounding stopped working
749
750## 2.0.8
751
752- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
753- IDE: Add drag-and-drop support for files and folders in chat
754- /context: Fix counting for thinking blocks
755- Improve message rendering for users with light themes on dark terminals
756- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
757
758## 2.0.5
759
760- IDE: Fix IME unintended message submission with Enter and Tab
761- IDE: Add "Open in Terminal" link in login screen
762- Fix unhandled OAuth expiration 401 API errors
763- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
764
765## 2.0.1
766
767- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
768- Various bug fixes and presentation improvements
769
770## 2.0.0
771
772- New native VS Code extension
773- Fresh coat of paint throughout the whole app
774- /rewind a conversation to undo code changes
775- /usage command to see plan limits
776- Tab to toggle thinking (sticky across sessions)
777- Ctrl-R to search history
778- Unshipped claude config command
779- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
780- SDK: The Claude Code SDK is now the Claude Agent SDK
781- Add subagents dynamically with `--agents` flag
782
783## 1.0.126
784
785- Enable /context command for Bedrock and Vertex
786- Add mTLS support for HTTP-based OpenTelemetry exporters
787
788## 1.0.124
789
790- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
791- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
792- No longer inform Claude of the list of allowed tools when permission is denied
793- Fixed security vulnerability in Bash tool permission checks
794- Improved VSCode extension performance for large files
795
796## 1.0.123
797
798- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
799- Fixed thinking mode triggering on negation phrases like "don't think"
800- Fixed rendering performance degradation during token streaming
801- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
802- Enhanced BashTool environment snapshot logging
803- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
804- Migrated --debug logging to a file, to enable easy tailing & filtering
805
806## 1.0.120
807
808- Fix input lag during typing, especially noticeable with large prompts
809- Improved VSCode extension command registry and sessions dialog user experience
810- Enhanced sessions dialog responsiveness and visual feedback
811- Fixed IDE compatibility issue by removing worktree support check
812- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
813
814## 1.0.119
815
816- Fix Windows issue where process visually freezes on entering interactive mode
817- Support dynamic headers for MCP servers via headersHelper configuration
818- Fix thinking mode not working in headless sessions
819- Fix slash commands now properly update allowed tools instead of replacing them
820
821## 1.0.117
822
823- Add Ctrl-R history search to recall previous commands like bash/zsh
824- Fix input lag while typing, especially on Windows
825- Add sed command to auto-allowed commands in acceptEdits mode
826- Fix Windows PATH comparison to be case-insensitive for drive letters
827- Add permissions management hint to /add-dir output
828
829## 1.0.115
830
831- Improve thinking mode display with enhanced visual effects
832- Type /t to temporarily disable thinking mode in your prompt
833- Improve path validation for glob and grep tools
834- Show condensed output for post-tool hooks to reduce visual clutter
835- Fix visual feedback when loading state completes
836- Improve UI consistency for permission request dialogs
837
838## 1.0.113
839
840- Deprecated piped input in interactive mode
841- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
842
843## 1.0.112
844
845- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
846- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
847- Hooks: Added systemMessage support for SessionEnd hooks
848- Added `spinnerTipsEnabled` setting to disable spinner tips
849- IDE: Various improvements and bug fixes
850
851## 1.0.111
852
853- /model now validates provided model names
854- Fixed Bash tool crashes caused by malformed shell syntax parsing
855
856## 1.0.110
857
858- /terminal-setup command now supports WezTerm
859- MCP: OAuth tokens now proactively refresh before expiration
860- Fixed reliability issues with background Bash processes
861
862## 1.0.109
863
864- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
865
866## 1.0.106
867
868- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
869
870## 1.0.97
871
872- Settings: /doctor now validates permission rule syntax and suggests corrections
873
874## 1.0.94
875
876- Vertex: add support for global endpoints for supported models
877- /memory command now allows direct editing of all imported memory files
878- SDK: Add custom tools as callbacks
879- Added /todos command to list current todo items
880
881## 1.0.93
882
883- Windows: Add alt + v shortcut for pasting images from clipboard
884- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
885
886## 1.0.90
887
888- Settings file changes take effect immediately - no restart required
889
890## 1.0.88
891
892- Fixed issue causing "OAuth authentication is currently not supported"
893- Status line input now includes `exceeds_200k_tokens`
894- Fixed incorrect usage tracking in /cost.
895- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
896- Bedrock: Updated default Sonnet model to Sonnet 4
897
898## 1.0.86
899
900- Added /context to help users self-serve debug context issues
901- SDK: Added UUID support for all SDK messages
902- SDK: Added `--replay-user-messages` to replay user messages back to stdout
903
904## 1.0.85
905
906- Status line input now includes session cost info
907- Hooks: Introduced SessionEnd hook
908
909## 1.0.84
910
911- Fix tool_use/tool_result id mismatch error when network is unstable
912- Fix Claude sometimes ignoring real-time steering when wrapping up a task
913- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
914- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
915
916## 1.0.83
917
918- @-mention: Support files with spaces in path
919- New shimmering spinner
920
921## 1.0.82
922
923- SDK: Add request cancellation support
924- SDK: New additionalDirectories option to search custom paths, improved slash command processing
925- Settings: Validation prevents invalid fields in .claude/settings.json files
926- MCP: Improve tool name consistency
927- Bash: Fix crash when Claude tries to automatically read large files
928
929## 1.0.81
930
931- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
932- Agents: Fix custom agent loading when agent files are unparsable
933
934## 1.0.80
935
936- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
937
938## 1.0.77
939
940- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
941- SDK: Add session support and permission denial tracking
942- Fix token limit errors in conversation summarization
943- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
944
945## 1.0.73
946
947- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
948- MCP: Press Esc to cancel OAuth authentication flows
949- Bash: Improved command validation and reduced false security warnings
950- UI: Enhanced spinner animations and status line visual hierarchy
951- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
952
953## 1.0.72
954
955- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
956
957## 1.0.71
958
959- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
960- Customizable status line: add your terminal prompt to Claude Code with /statusline
961
962## 1.0.70
963
964- Performance: Optimized message rendering for better performance with large contexts
965- Windows: Fixed native file search, ripgrep, and subagent functionality
966- Added support for @-mentions in slash command arguments
967
968## 1.0.69
969
970- Upgraded Opus to version 4.1
971
972## 1.0.68
973
974- Fix incorrect model names being used for certain commands like `/pr-comments`
975- 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.
976- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
977- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
978- SDK: Added canUseTool callback support for tool confirmation
979- Added `disableAllHooks` setting
980- Improved file suggestions performance in large repos
981
982## 1.0.65
983
984- IDE: Fixed connection stability issues and error handling for diagnostics
985- Windows: Fixed shell environment setup for users without .bashrc files
986
987## 1.0.64
988
989- Agents: Added model customization support - you can now specify which model an agent should use
990- Agents: Fixed unintended access to the recursive agent tool
991- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
992- SDK: Fixed user input tracking across multi-turn conversations
993- Added hidden files to file search and @-mention suggestions
994
995## 1.0.63
996
997- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
998
999## 1.0.62
1000
1001- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
1002- Hooks: Added SessionStart hook for new session initialization
1003- /add-dir command now supports typeahead for directory paths
1004- Improved network connectivity check reliability
1005
1006## 1.0.61
1007
1008- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
1009- Settings: Added `--settings` flag to load settings from a JSON file
1010- Settings: Fixed resolution of settings files paths that are symlinks
1011- OTEL: Fixed reporting of wrong organization after authentication changes
1012- Slash commands: Fixed permissions checking for allowed-tools with Bash
1013- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
1014- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
1015- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
1016
1017## 1.0.60
1018
1019- You can now create custom subagents for specialized tasks! Run /agents to get started
1020
1021## 1.0.59
1022
1023- SDK: Added tool confirmation support with canUseTool callback
1024- SDK: Allow specifying env for spawned process
1025- Hooks: Exposed PermissionDecision to hooks (including "ask")
1026- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
1027- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
1028
1029## 1.0.58
1030
1031- Added support for reading PDFs
1032- MCP: Improved server health status display in 'claude mcp list'
1033- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
1034
1035## 1.0.57
1036
1037- Added support for specifying a model in slash commands
1038- Improved permission messages to help Claude understand allowed tools
1039- Fix: Remove trailing newlines from bash output in terminal wrapping
1040
1041## 1.0.56
1042
1043- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
1044- Fixes for WSL IDE detection
1045- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
1046
1047## 1.0.55
1048
1049- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
1050- Windows: fixed Ctrl+Z crash
1051- SDK: Added ability to capture error logging
1052- Add --system-prompt-file option to override system prompt in print mode
1053
1054## 1.0.54
1055
1056- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
1057- Custom slash commands: Added argument-hint to frontmatter
1058- Windows: OAuth uses port 45454 and properly constructs browser URL
1059- Windows: mode switching now uses alt + m, and plan mode renders properly
1060- Shell: Switch to in-memory shell snapshot to fix file-related errors
1061
1062## 1.0.53
1063
1064- Updated @-mention file truncation from 100 lines to 2000 lines
1065- 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).
1066
1067## 1.0.52
1068
1069- Added support for MCP server instructions
1070
1071## 1.0.51
1072
1073- Added support for native Windows (requires Git for Windows)
1074- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
1075- Settings: /doctor can now help you identify and fix invalid setting files
1076- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
1077- Increased auto-compact warning threshold from 60% to 80%
1078- Fixed an issue with handling user directories with spaces for shell snapshots
1079- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
1080- Custom slash commands: Fixed user-level commands in subdirectories
1081- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
1082
1083## 1.0.48
1084
1085- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
1086- Added progress messages to Bash tool based on the last 5 lines of command output
1087- Added expanding variables support for MCP server configuration
1088- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
1089- Improved IDE extension path handling when Claude Code runs in WSL
1090- Hooks: Added a PreCompact hook
1091- Vim mode: Added c, f/F, t/T
1092
1093## 1.0.45
1094
1095- Redesigned Search (Grep) tool with new tool input parameters and features
1096- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
1097- Fixed config file corruption issue by enforcing atomic writes
1098- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
1099- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
1100- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
1101
1102## 1.0.44
1103
1104- New /export command lets you quickly export a conversation for sharing
1105- MCP: resource_link tool results are now supported
1106- MCP: tool annotations and tool titles now display in /mcp view
1107- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
1108
1109## 1.0.43
1110
1111- Fixed a bug where the theme selector was saving excessively
1112- Hooks: Added EPIPE system error handling
1113
1114## 1.0.42
1115
1116- Added tilde (`~`) expansion support to `/add-dir` command
1117
1118## 1.0.41
1119
1120- Hooks: Split Stop hook triggering into Stop and SubagentStop
1121- Hooks: Enabled optional timeout configuration for each command
1122- Hooks: Added "hook_event_name" to hook input
1123- Fixed a bug where MCP tools would display twice in tool list
1124- New tool parameters JSON for Bash tool in `tool_decision` event
1125
1126## 1.0.40
1127
1128- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1129
1130## 1.0.39
1131
1132- New Active Time metric in OpenTelemetry logging
1133
1134## 1.0.38
1135
1136- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1137
1138## 1.0.37
1139
1140- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1141
1142## 1.0.36
1143
1144- Web search now takes today's date into context
1145- Fixed a bug where stdio MCP servers were not terminating properly on exit
1146
1147## 1.0.35
1148
1149- Added support for MCP OAuth Authorization Server discovery
1150
1151## 1.0.34
1152
1153- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1154
1155## 1.0.33
1156
1157- Improved logging functionality with session ID support
1158- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1159- Improvements to plan mode
1160
1161## 1.0.32
1162
1163- Updated loopback config for litellm
1164- Added forceLoginMethod setting to bypass login selection screen
1165
1166## 1.0.31
1167
1168- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1169
1170## 1.0.30
1171
1172- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1173- Improved file path autocomplete with filename matching
1174- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1175- Enhanced jq regex support for complex filters with pipes and select
1176
1177## 1.0.29
1178
1179- Improved CJK character support in cursor navigation and rendering
1180
1181## 1.0.28
1182
1183- Slash commands: Fix selector display during history navigation
1184- Resizes images before upload to prevent API size limit errors
1185- Added XDG_CONFIG_HOME support to configuration directory
1186- Performance optimizations for memory usage
1187- New attributes (terminal.type, language) in OpenTelemetry logging
1188
1189## 1.0.27
1190
1191- Streamable HTTP MCP servers are now supported
1192- Remote MCP servers (SSE and HTTP) now support OAuth
1193- MCP resources can now be @-mentioned
1194- /resume slash command to switch conversations within Claude Code
1195
1196## 1.0.25
1197
1198- Slash commands: moved "project" and "user" prefixes to descriptions
1199- Slash commands: improved reliability for command discovery
1200- Improved support for Ghostty
1201- Improved web search reliability
1202
1203## 1.0.24
1204
1205- Improved /mcp output
1206- Fixed a bug where settings arrays got overwritten instead of merged
1207
1208## 1.0.23
1209
1210- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1211- Released Python SDK: pip install claude-code-sdk to get started
1212
1213## 1.0.22
1214
1215- SDK: Renamed `total_cost` to `total_cost_usd`
1216
1217## 1.0.21
1218
1219- Improved editing of files with tab-based indentation
1220- Fix for tool_use without matching tool_result errors
1221- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1222
1223## 1.0.18
1224
1225- Added --add-dir CLI argument for specifying additional working directories
1226- Added streaming input support without require -p flag
1227- Improved startup performance and session storage performance
1228- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1229- Added detailed MCP server tools display (/mcp)
1230- MCP authentication and permission improvements
1231- Added auto-reconnection for MCP SSE connections on disconnect
1232- Fixed issue where pasted content was lost when dialogs appeared
1233
1234## 1.0.17
1235
1236- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1237- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1238- MCP server list UI improvements
1239- Update Claude Code process title to display "claude" instead of "node"
1240
1241## 1.0.11
1242
1243- Claude Code can now also be used with a Claude Pro subscription
1244- Added /upgrade for smoother switching to Claude Max plans
1245- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1246- Improved shell configuration error handling
1247- Improved todo list handling during compaction
1248
1249## 1.0.10
1250
1251- Added markdown table support
1252- Improved streaming performance
1253
1254## 1.0.8
1255
1256- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1257- Increased default otel interval from 1s -> 5s
1258- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1259- Fixed a regression where search tools unnecessarily asked for permissions
1260- Added support for triggering thinking non-English languages
1261- Improved compacting UI
1262
1263## 1.0.7
1264
1265- Renamed /allowed-tools -> /permissions
1266- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1267- Deprecated claude config commands in favor of editing settings.json
1268- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1269- Improved error handling for /install-github-app
1270- Bugfixes, UI polish, and tool reliability improvements
1271
1272## 1.0.6
1273
1274- Improved edit reliability for tab-indented files
1275- Respect CLAUDE_CONFIG_DIR everywhere
1276- Reduced unnecessary tool permission prompts
1277- Added support for symlinks in @file typeahead
1278- Bugfixes, UI polish, and tool reliability improvements
1279
1280## 1.0.4
1281
1282- Fixed a bug where MCP tool errors weren't being parsed correctly
1283
1284## 1.0.1
1285
1286- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1287- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1288- Updated documentation links and OAuth process descriptions
1289
1290## 1.0.0
1291
1292- Claude Code is now generally available
1293- Introducing Sonnet 4 and Opus 4 models
1294
1295## 0.2.125
1296
1297- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1298- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1299
1300## 0.2.117
1301
1302- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1303- Introduced settings.cleanupPeriodDays
1304- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1305- Introduced --debug mode
1306
1307## 0.2.108
1308
1309- You can now send messages to Claude while it works to steer Claude in real-time
1310- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1311- Fixed a bug where thinking was not working in -p mode
1312- Fixed a regression in /cost reporting
1313- Deprecated MCP wizard interface in favor of other MCP commands
1314- Lots of other bugfixes and improvements
1315
1316## 0.2.107
1317
1318- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1319
1320## 0.2.106
1321
1322- MCP SSE server configs can now specify custom headers
1323- Fixed a bug where MCP permission prompt didn't always show correctly
1324
1325## 0.2.105
1326
1327- Claude can now search the web
1328- Moved system & account status to /status
1329- Added word movement keybindings for Vim
1330- Improved latency for startup, todo tool, and file edits
1331
1332## 0.2.102
1333
1334- Improved thinking triggering reliability
1335- Improved @mention reliability for images and folders
1336- You can now paste multiple large chunks into one prompt
1337
1338## 0.2.100
1339
1340- Fixed a crash caused by a stack overflow error
1341- Made db storage optional; missing db support disables --continue and --resume
1342
1343## 0.2.98
1344
1345- Fixed an issue where auto-compact was running twice
1346
1347## 0.2.96
1348
1349- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1350
1351## 0.2.93
1352
1353- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1354- Claude now has access to a Todo list that helps it stay on track and be more organized
1355
1356## 0.2.82
1357
1358- Added support for --disallowedTools
1359- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1360
1361## 0.2.75
1362
1363- Hit Enter to queue up additional messages while Claude is working
1364- Drag in or copy/paste image files directly into the prompt
1365- @-mention files to directly add them to context
1366- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1367- Improved performance for filename auto-complete
1368
1369## 0.2.74
1370
1371- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1372- Task tool can now perform writes and run bash commands
1373
1374## 0.2.72
1375
1376- Updated spinner to indicate tokens loaded and tool usage
1377
1378## 0.2.70
1379
1380- Network commands like curl are now available for Claude to use
1381- Claude can now run multiple web queries in parallel
1382- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1383
1384## 0.2.69
1385
1386- Fixed UI glitches with improved Select component behavior
1387- Enhanced terminal output display with better text truncation logic
1388
1389## 0.2.67
1390
1391- Shared project permission rules can be saved in .claude/settings.json
1392
1393## 0.2.66
1394
1395- Print mode (-p) now supports streaming output via --output-format=stream-json
1396- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1397
1398## 0.2.63
1399
1400- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1401
1402## 0.2.61
1403
1404- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1405- Enhanced image detection for more reliable clipboard paste functionality
1406- Fixed an issue where ESC key could crash the conversation history selector
1407
1408## 0.2.59
1409
1410- Copy+paste images directly into your prompt
1411- Improved progress indicators for bash and fetch tools
1412- Bugfixes for non-interactive mode (-p)
1413
1414## 0.2.54
1415
1416- Quickly add to Memory by starting your message with '#'
1417- Press ctrl+r to see full output for long tool results
1418- Added support for MCP SSE transport
1419
1420## 0.2.53
1421
1422- New web fetch tool lets Claude view URLs that you paste in
1423- Fixed a bug with JPEG detection
1424
1425## 0.2.50
1426
1427- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1428
1429## 0.2.49
1430
1431- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1432
1433## 0.2.47
1434
1435- Press Tab to auto-complete file and folder names
1436- Press Shift + Tab to toggle auto-accept for file edits
1437- Automatic conversation compaction for infinite conversation length (toggle with /config)
1438
1439## 0.2.44
1440
1441- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1442
1443## 0.2.41
1444
1445- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1446- MCP server startup no longer blocks the app from starting up
1447
1448## 0.2.37
1449
1450- New /release-notes command lets you view release notes at any time
1451- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1452
1453## 0.2.36
1454
1455- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1456- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1457
1458## 0.2.34
1459
1460- Vim bindings for text input - enable with /vim or /config
1461
1462## 0.2.32
1463
1464- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1465- Fix for some PersistentShell issues
1466
1467## 0.2.31
1468
1469- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1470- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1471
1472## 0.2.30
1473
1474- Added ANSI color theme for better terminal compatibility
1475- Fixed issue where slash command arguments weren't being sent properly
1476- (Mac-only) API keys are now stored in macOS Keychain
1477
1478## 0.2.26
1479
1480- New /approved-tools command for managing tool permissions
1481- Word-level diff display for improved code readability
1482- Fuzzy matching for slash commands
1483
1484## 0.2.21
1485
1486- Fuzzy matching for /commands