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 74cc597 · on Jan 17, 2026 History
1,424 lines · 70.1 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.12
4
5- Fixed message rendering bug
6
7## 2.1.11
8
9- Fixed excessive MCP connection requests for HTTP/SSE transports
10
11## 2.1.10
12
13- Added new `Setup` hook event that can be triggered via `--init`, `--init-only`, or `--maintenance` CLI flags for repository setup and maintenance operations
14- Added keyboard shortcut 'c' to copy OAuth URL when browser doesn't open automatically during login
15- Fixed a crash when running bash commands containing heredocs with JavaScript template literals like `${index + 1}`
16- Improved startup to capture keystrokes typed before the REPL is fully ready
17- Improved file suggestions to show as removable attachments instead of inserting text when accepted
18- [VSCode] Added install count display to plugin listings
19- [VSCode] Added trust warning when installing plugins
20
21## 2.1.9
22
23- Added `auto:N` syntax for configuring the MCP tool search auto-enable threshold, where N is the context window percentage (0-100)
24- Added `plansDirectory` setting to customize where plan files are stored
25- Added external editor support (Ctrl+G) in AskUserQuestion "Other" input field
26- Added session URL attribution to commits and PRs created from web sessions
27- Added support for `PreToolUse` hooks to return `additionalContext` to the model
28- Added `${CLAUDE_SESSION_ID}` string substitution for skills to access the current session ID
29- Fixed long sessions with parallel tool calls failing with an API error about orphan tool_result blocks
30- Fixed MCP server reconnection hanging when cached connection promise never resolves
31- Fixed Ctrl+Z suspend not working in terminals using Kitty keyboard protocol (Ghostty, iTerm2, kitty, WezTerm)
32
33## 2.1.7
34
35- Added `showTurnDuration` setting to hide turn duration messages (e.g., "Cooked for 1m 6s")
36- Added ability to provide feedback when accepting permission prompts
37- Added inline display of agent's final response in task notifications, making it easier to see results without reading the full transcript file
38- Fixed security vulnerability where wildcard permission rules could match compound commands containing shell operators
39- Fixed false "file modified" errors on Windows when cloud sync tools, antivirus scanners, or Git touch file timestamps without changing content
40- Fixed orphaned tool_result errors when sibling tools fail during streaming execution
41- 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)
42- Fixed spinner briefly flashing when running local slash commands like `/model` or `/theme`
43- Fixed terminal title animation jitter by using fixed-width braille characters
44- Fixed plugins with git submodules not being fully initialized when installed
45- Fixed bash commands failing on Windows when temp directory paths contained characters like `t` or `n` that were misinterpreted as escape sequences
46- Improved typing responsiveness by reducing memory allocation overhead in terminal rendering
47- 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.
48- Changed OAuth and API Console URLs from console.anthropic.com to platform.claude.com
49- [VSCode] Fixed `claudeProcessWrapper` setting passing the wrapper path instead of the Claude binary path
50
51## 2.1.6
52
53- Added search functionality to `/config` command for quickly filtering settings
54- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
55- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
56- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
57- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
58- Added an error display when the editor fails during Ctrl+G
59- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
60- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
61- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
62- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
63- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
64- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
65- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
66- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
67- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
68- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
69- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
70- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
71- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
72- Improved @ autocomplete with icons for different suggestion types and single-line formatting
73- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
74- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
75- Changed terminal title to "Claude Code" on startup for better window identification
76- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
77- [VSCode] Fixed usage indicator not updating after manual compact
78
79## 2.1.5
80
81- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
82
83## 2.1.4
84
85- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
86- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
87
88## 2.1.3
89
90- Merged slash commands and skills, simplifying the mental model with no change in behavior
91- Added release channel (`stable` or `latest`) toggle to `/config`
92- 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
93- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
94- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
95- Fixed mismatch between background task count in status bar and items shown in tasks dialog
96- Fixed sub-agents using the wrong model during conversation compaction
97- Fixed web search in sub-agents using incorrect model
98- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
99- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
100- Improved slash command suggestion readability by truncating long descriptions to 2 lines
101- Changed tool hook execution timeout from 60 seconds to 10 minutes
102- [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)
103
104## 2.1.2
105
106- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
107- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
108- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
109- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
110- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
111- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
112- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
113- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
114- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
115- Fixed updates incorrectly claiming another installation is in progress
116- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
117- Fixed remote session URL and teleport being broken when using `/tasks` command
118- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
119- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
120- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
121- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
122- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
123- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
124- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
125- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
126- [SDK] Changed minimum zod peer dependency to ^4.0.0
127- [VSCode] Fixed usage display not updating after manual compact
128
129## 2.1.0
130
131- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
132- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
133- Added support for `agent` field in skills to specify agent type for execution
134- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
135- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
136- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
137- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
138- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
139- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
140- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
141- Fixed Esc key with queued prompts to only move them to input without canceling the running task
142- Reduced permission prompts for complex bash commands
143- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
144- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
145- Fixed plugin path resolution for file-based marketplace sources
146- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
147- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
148- Fixed Claude in Chrome support for WSL environments
149- Fixed Windows native installer silently failing when executable creation fails
150- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
151- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
152- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
153- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
154- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
155- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
156- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
157- Added hooks support for skill and slash command frontmatter
158- 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
159- Added `/plan` command shortcut to enable plan mode directly from the prompt
160- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
161- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
162- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
163- Added support for `once: true` config for hooks
164- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
165- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
166- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
167- Added left/right arrow key navigation for cycling through tabs in dialogs
168- Added real-time thinking block display in Ctrl+O transcript mode
169- Added filepath to full output in background bash task details dialog
170- Added Skills as a separate category in the context visualization
171- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
172- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
173- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
174- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
175- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
176- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
177- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
178- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
179- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
180- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
181- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
182- Fixed commands with `$()` command substitution failing with parse errors
183- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
184- 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)
185- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
186- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
187- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
188- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
189- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
190- Fixed subagents sometimes not inheriting the parent's model by default
191- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
192- Fixed duplicate Bash commands appearing in permission request option labels
193- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
194- Fixed background task completion notifications to appear proactively with bullet point
195- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
196- Fixed cursor disappearing after dismissing permission dialogs
197- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
198- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
199- Fixed images being silently dropped when queueing messages while backgrounding a task
200- Fixed large pasted images failing with "Image was too large" error
201- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
202- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
203- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
204- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
205- Fixed spinner token counter to properly accumulate tokens from subagents during execution
206- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
207- Fixed race condition where LSP tool could return "no server available" during startup
208- Fixed feedback submission hanging indefinitely when network requests timeout
209- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
210- Fixed hook success message showing trailing colon when hook has no output
211- Multiple optimizations to improve startup performance
212- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
213- Improved performance when reading Jupyter notebooks with many cells
214- Improved reliability for piped input like `cat refactor.md | claude`
215- Improved reliability for AskQuestion tool
216- Improved sed in-place edit commands to render as file edits with diff preview
217- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
218- Improved compaction reliability
219- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
220- Improved skills to show progress while executing, displaying tool uses as they happen
221- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
222- Improved skill suggestions to prioritize recently and frequently used skills
223- Improved spinner feedback when waiting for the first response token
224- Improved token count display in spinner to include tokens from background agents
225- Improved incremental output for async agents to give the main thread more control and visibility
226- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
227- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
228- Improved macOS screenshot paste reliability with TIFF format support
229- Improved `/stats` output
230- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
231- Changed "Interrupted" message color from red to grey for a less alarming appearance
232- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
233- Removed underline styling from image reference links
234- [SDK] Changed minimum zod peer dependency to ^4.0.0
235- [VSCode] Added currently selected model name to the context menu
236- [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")
237- [VSCode] Fixed paragraph breaks not rendering in markdown content
238- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
239- [Windows] Fixed issue with improper rendering
240
241## 2.0.76
242
243- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
244
245## 2.0.75
246
247- Minor bugfixes
248
249## 2.0.74
250
251- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
252- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
253- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
254- Added syntax highlighting info to theme picker
255- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
256- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
257- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
258- Fixed a potential crash when syntax highlighting isn't initialized correctly
259- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
260- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
261- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
262- [Windows] Fixed issue with improper rendering
263- [VSCode] Added gift tag pictogram for year-end promotion message
264
265## 2.0.73
266
267- Added clickable `[Image #N]` links that open attached images in the default viewer
268- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
269- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
270- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
271- Fixed slow input history cycling and race condition that could overwrite text after message submission
272- Improved `/theme` command to open theme picker directly
273- Improved theme picker UI
274- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
275- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
276
277## 2.0.72
278
279- 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
280- Reduced terminal flickering
281- Added scannable QR code to mobile app tip for quick app downloads
282- Added loading indicator when resuming conversations for better feedback
283- Fixed `/context` command not respecting custom system prompts in non-interactive mode
284- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
285- Improved @ mention file suggestion speed (~3x faster in git repositories)
286- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
287- Improved settings validation errors to be more prominent
288- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
289
290## 2.0.71
291
292- Added /config toggle to enable/disable prompt suggestions
293- Added `/settings` as an alias for the `/config` command
294- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
295- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
296- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
297- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
298- New syntax highlighting engine for native build
299
300## 2.0.70
301
302- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
303- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
304- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
305- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
306- Fixed input being cleared when processing queued commands while the user was typing
307- Fixed prompt suggestions replacing typed input when pressing Tab
308- Fixed diff view not updating when terminal is resized
309- Improved memory usage by 3x for large conversations
310- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
311- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
312- Fix thinking mode toggle in /config not persisting correctly
313- Improve UI for file creation permission dialog
314
315## 2.0.69
316
317- Minor bugfixes
318
319## 2.0.68
320
321- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
322- Fixed a bug where disallowed MCP tools were visible to the model
323- Fixed an issue where steering messages could be lost while a subagent is working
324- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
325- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
326- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
327
328## 2.0.67
329
330- Thinking mode is now enabled by default for Opus 4.5
331- Thinking mode configuration has moved to /config
332- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
333- Show reason why autoupdater is disabled in `/doctor`
334- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
335- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
336- Fixed scroll position resetting after deleting a permission rule in `/permissions`
337- 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
338- Fixed `claude install --force` not bypassing stale lock files
339- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
340- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
341
342## 2.0.65
343
344- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
345- Added context window information to status line input
346- Added `fileSuggestion` setting for custom `@` file search commands
347- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
348- Fixed prompt not being saved to history when aborting a query with Escape
349- Fixed Read tool image handling to identify format from bytes instead of file extension
350
351## 2.0.64
352
353- Made auto-compacting instant
354- Agents and bash commands can run asynchronously and send messages to wake up the main agent
355- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
356- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
357- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
358- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
359- Fixed auto-loading .env when using native installer
360- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
361- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
362- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
363- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
364- Bedrock: Improve efficiency of token counting
365- Bedrock: Add support for `aws login` AWS Management Console credentials
366- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
367
368## 2.0.62
369
370- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
371- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
372- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
373- Fixed slash command selection not working when multiple commands share the same name
374- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
375- Fixed running versions getting removed because lock file incorrectly going stale
376- Fixed IDE diff tab not closing when rejecting file changes
377
378## 2.0.61
379
380- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
381
382## 2.0.60
383
384- Added background agent support. Agents run in the background while you work
385- Added --disable-slash-commands CLI flag to disable all slash commands
386- Added model name to "Co-Authored-By" commit messages
387- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
388- Updated Fetch to skip summarization for pre-approved websites
389- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
390
391## 2.0.59
392
393- Added --agent CLI flag to override the agent setting for the current session
394- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
395- VS Code: Fixed .claude.json config file being read from incorrect location
396
397## 2.0.58
398
399- Pro users now have access to Opus 4.5 as part of their subscription!
400- Fixed timer duration showing "11m 60s" instead of "12m 0s"
401- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
402
403## 2.0.57
404
405- Added feedback input when rejecting plans, allowing users to tell Claude what to change
406- VSCode: Added streaming message support for real-time response display
407
408## 2.0.56
409
410- Added setting to enable/disable terminal progress bar (OSC 9;4)
411- 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.
412
413## 2.0.55
414
415- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
416- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
417- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
418- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
419
420## 2.0.54
421
422- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
423- Fix issue with excessive iTerm notifications
424
425## 2.0.52
426
427- Fixed duplicate message display when starting Claude with a command line argument
428- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
429- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
430- Permit some uses of `$!` in bash commands
431
432## 2.0.51
433
434- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
435- Introducing Claude Code for Desktop: https://claude.com/download
436- 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
437- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
438- Plan Mode now builds more precise plans and executes more thoroughly
439- Usage limit notifications now easier to understand
440- Switched `/usage` back to "% used"
441- Fixed handling of thinking errors
442- Fixed performance regression
443
444## 2.0.50
445
446- Fixed bug preventing calling MCP tools that have nested references in their input schemas
447- Silenced a noisy but harmless error during upgrades
448- Improved ultrathink text display
449- Improved clarity of 5-hour session limit warning message
450
451## 2.0.49
452
453- Added readline-style ctrl-y for pasting deleted text
454- Improved clarity of usage limit warning message
455- Fixed handling of subagent permissions
456
457## 2.0.47
458
459- Improved error messages and validation for `claude --teleport`
460- Improved error handling in `/usage`
461- Fixed race condition with history entry not getting logged at exit
462- Fixed Vertex AI configuration not being applied from `settings.json`
463
464## 2.0.46
465
466- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
467
468## 2.0.45
469
470- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
471- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
472- Send background tasks to Claude Code on the web by starting a message with `&`
473
474## 2.0.43
475
476- Added `permissionMode` field for custom agents
477- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
478- Added skills frontmatter field to declare skills to auto-load for subagents
479- Added the `SubagentStart` hook event
480- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
481- Fixed duplicate rendering of some messages in the UI
482- Fixed some visual flickers
483- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
484
485## 2.0.42
486
487- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
488
489## 2.0.41
490
491- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
492- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
493- Fixed incorrect labeling of user settings vs project settings in command descriptions
494- Fixed crash when plugin command hooks timeout during execution
495- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
496- Fixed broken security documentation links in trust dialogs and onboarding
497- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
498- ctrl-r history search landing on a slash command no longer cancels the search
499- SDK: Support custom timeouts for hooks
500- Allow more safe git commands to run without approval
501- Plugins: Added support for sharing and installing output styles
502- Teleporting a session from web will automatically set the upstream branch
503
504## 2.0.37
505
506- Fixed how idleness is computed for notifications
507- Hooks: Added matcher values for Notification hook events
508- Output Styles: Added `keep-coding-instructions` option to frontmatter
509
510## 2.0.36
511
512- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
513- Fixed queued messages being incorrectly executed as bash commands
514- Fixed input being lost when typing while a queued message is processed
515
516## 2.0.35
517
518- Improve fuzzy search results when searching commands
519- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
520- 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
521- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
522- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
523
524## 2.0.34
525
526- VSCode Extension: Added setting to configure the initial permission mode for new conversations
527- Improved file path suggestion performance with native Rust-based fuzzy finder
528- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
529- Fixed memory crash when reading or writing large files (especially base64-encoded images)
530
531## 2.0.33
532
533- Native binary installs now launch quicker.
534- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
535- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
536
537## 2.0.32
538
539- Un-deprecate output styles based on community feedback
540- Added `companyAnnouncements` setting for displaying announcements on startup
541- Fixed hook progress messages not updating correctly during PostToolUse hook execution
542
543## 2.0.31
544
545- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
546- Vertex: add support for Web Search on supported models
547- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
548- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
549- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
550- Fixed plugin uninstall not removing plugins
551
552## 2.0.30
553
554- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
555- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
556- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
557- Added prompt-based stop hooks
558- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
559- Enabled SSE MCP servers on native build
560- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
561- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
562- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
563- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
564- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
565- Fixed bug that saved session permissions to local settings
566- Fixed MCP tools not being available to sub-agents
567- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
568- Fixed delay when navigating through typeahead suggestions with arrow keys
569- VSCode: Restored selection indicator in input footer showing current file or code selection status
570
571## 2.0.28
572
573- Plan mode: introduced new Plan subagent
574- Subagents: claude can now choose to resume subagents
575- Subagents: claude can dynamically choose the model used by its subagents
576- SDK: added --max-budget-usd flag
577- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
578- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
579- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
580- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
581- Various other bug fixes
582
583## 2.0.27
584
585- New UI for permission prompts
586- Added current branch filtering and search to session resume screen for easier navigation
587- Fixed directory @-mention causing "No assistant message found" error
588- VSCode Extension: Add config setting to include .gitignored files in file searches
589- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
590
591## 2.0.25
592
593- 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
594
595## 2.0.24
596
597- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
598- Claude Code Web: Support for Web -> CLI teleport
599- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
600- Bedrock: Display awsAuthRefresh output when auth is required
601
602## 2.0.22
603
604- Fixed content layout shift when scrolling through slash commands
605- IDE: Add toggle to enable/disable thinking.
606- Fix bug causing duplicate permission prompts with parallel tool calls
607- Add support for enterprise managed MCP allowlist and denylist
608
609## 2.0.21
610
611- Support MCP `structuredContent` field in tool responses
612- Added an interactive question tool
613- Claude will now ask you questions more often in plan mode
614- Added Haiku 4.5 as a model option for Pro users
615- Fixed an issue where queued commands don't have access to previous messages' output
616
617## 2.0.20
618
619- Added support for Claude Skills
620
621## 2.0.19
622
623- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
624- Fixed a bug where Haiku was unnecessarily called in print mode
625
626## 2.0.17
627
628- Added Haiku 4.5 to model selector!
629- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
630- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
631- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
632- OTEL: support HTTP_PROXY and HTTPS_PROXY
633- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
634
635## 2.0.15
636
637- Fixed bug with resuming where previously created files needed to be read again before writing
638- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
639
640## 2.0.14
641
642- Fix @-mentioning MCP servers to toggle them on/off
643- Improve permission checks for bash with inline env vars
644- Fix ultrathink + thinking toggle
645- Reduce unnecessary logins
646- Document --system-prompt
647- Several improvements to rendering
648- Plugins UI polish
649
650## 2.0.13
651
652- Fixed `/plugin` not working on native build
653
654## 2.0.12
655
656- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
657- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
658- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
659- `/plugin validate` command for validating plugin structure and configuration
660- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
661- Plugin documentation available at https://code.claude.com/docs/en/plugins
662- Comprehensive error messages and diagnostics via `/doctor` command
663- Avoid flickering in `/model` selector
664- Improvements to `/help`
665- Avoid mentioning hooks in `/resume` summaries
666- Changes to the "verbose" setting in `/config` now persist across sessions
667
668## 2.0.11
669
670- Reduced system prompt size by 1.4k tokens
671- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
672- Fixed Opus fallback rate limit errors appearing incorrectly
673- Fixed /add-dir command selecting wrong default tab
674
675## 2.0.10
676
677- Rewrote terminal renderer for buttery smooth UI
678- Enable/disable MCP servers by @mentioning, or in /mcp
679- Added tab completion for shell commands in bash mode
680- PreToolUse hooks can now modify tool inputs
681- Press Ctrl-G to edit your prompt in your system's configured text editor
682- Fixes for bash permission checks with environment variables in the command
683
684## 2.0.9
685
686- Fix regression where bash backgrounding stopped working
687
688## 2.0.8
689
690- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
691- IDE: Add drag-and-drop support for files and folders in chat
692- /context: Fix counting for thinking blocks
693- Improve message rendering for users with light themes on dark terminals
694- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
695
696## 2.0.5
697
698- IDE: Fix IME unintended message submission with Enter and Tab
699- IDE: Add "Open in Terminal" link in login screen
700- Fix unhandled OAuth expiration 401 API errors
701- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
702
703## 2.0.1
704
705- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
706- Various bug fixes and presentation improvements
707
708## 2.0.0
709
710- New native VS Code extension
711- Fresh coat of paint throughout the whole app
712- /rewind a conversation to undo code changes
713- /usage command to see plan limits
714- Tab to toggle thinking (sticky across sessions)
715- Ctrl-R to search history
716- Unshipped claude config command
717- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
718- SDK: The Claude Code SDK is now the Claude Agent SDK
719- Add subagents dynamically with `--agents` flag
720
721## 1.0.126
722
723- Enable /context command for Bedrock and Vertex
724- Add mTLS support for HTTP-based OpenTelemetry exporters
725
726## 1.0.124
727
728- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
729- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
730- No longer inform Claude of the list of allowed tools when permission is denied
731- Fixed security vulnerability in Bash tool permission checks
732- Improved VSCode extension performance for large files
733
734## 1.0.123
735
736- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
737- Fixed thinking mode triggering on negation phrases like "don't think"
738- Fixed rendering performance degradation during token streaming
739- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
740- Enhanced BashTool environment snapshot logging
741- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
742- Migrated --debug logging to a file, to enable easy tailing & filtering
743
744## 1.0.120
745
746- Fix input lag during typing, especially noticeable with large prompts
747- Improved VSCode extension command registry and sessions dialog user experience
748- Enhanced sessions dialog responsiveness and visual feedback
749- Fixed IDE compatibility issue by removing worktree support check
750- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
751
752## 1.0.119
753
754- Fix Windows issue where process visually freezes on entering interactive mode
755- Support dynamic headers for MCP servers via headersHelper configuration
756- Fix thinking mode not working in headless sessions
757- Fix slash commands now properly update allowed tools instead of replacing them
758
759## 1.0.117
760
761- Add Ctrl-R history search to recall previous commands like bash/zsh
762- Fix input lag while typing, especially on Windows
763- Add sed command to auto-allowed commands in acceptEdits mode
764- Fix Windows PATH comparison to be case-insensitive for drive letters
765- Add permissions management hint to /add-dir output
766
767## 1.0.115
768
769- Improve thinking mode display with enhanced visual effects
770- Type /t to temporarily disable thinking mode in your prompt
771- Improve path validation for glob and grep tools
772- Show condensed output for post-tool hooks to reduce visual clutter
773- Fix visual feedback when loading state completes
774- Improve UI consistency for permission request dialogs
775
776## 1.0.113
777
778- Deprecated piped input in interactive mode
779- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
780
781## 1.0.112
782
783- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
784- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
785- Hooks: Added systemMessage support for SessionEnd hooks
786- Added `spinnerTipsEnabled` setting to disable spinner tips
787- IDE: Various improvements and bug fixes
788
789## 1.0.111
790
791- /model now validates provided model names
792- Fixed Bash tool crashes caused by malformed shell syntax parsing
793
794## 1.0.110
795
796- /terminal-setup command now supports WezTerm
797- MCP: OAuth tokens now proactively refresh before expiration
798- Fixed reliability issues with background Bash processes
799
800## 1.0.109
801
802- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
803
804## 1.0.106
805
806- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
807
808## 1.0.97
809
810- Settings: /doctor now validates permission rule syntax and suggests corrections
811
812## 1.0.94
813
814- Vertex: add support for global endpoints for supported models
815- /memory command now allows direct editing of all imported memory files
816- SDK: Add custom tools as callbacks
817- Added /todos command to list current todo items
818
819## 1.0.93
820
821- Windows: Add alt + v shortcut for pasting images from clipboard
822- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
823
824## 1.0.90
825
826- Settings file changes take effect immediately - no restart required
827
828## 1.0.88
829
830- Fixed issue causing "OAuth authentication is currently not supported"
831- Status line input now includes `exceeds_200k_tokens`
832- Fixed incorrect usage tracking in /cost.
833- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
834- Bedrock: Updated default Sonnet model to Sonnet 4
835
836## 1.0.86
837
838- Added /context to help users self-serve debug context issues
839- SDK: Added UUID support for all SDK messages
840- SDK: Added `--replay-user-messages` to replay user messages back to stdout
841
842## 1.0.85
843
844- Status line input now includes session cost info
845- Hooks: Introduced SessionEnd hook
846
847## 1.0.84
848
849- Fix tool_use/tool_result id mismatch error when network is unstable
850- Fix Claude sometimes ignoring real-time steering when wrapping up a task
851- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
852- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
853
854## 1.0.83
855
856- @-mention: Support files with spaces in path
857- New shimmering spinner
858
859## 1.0.82
860
861- SDK: Add request cancellation support
862- SDK: New additionalDirectories option to search custom paths, improved slash command processing
863- Settings: Validation prevents invalid fields in .claude/settings.json files
864- MCP: Improve tool name consistency
865- Bash: Fix crash when Claude tries to automatically read large files
866
867## 1.0.81
868
869- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
870- Agents: Fix custom agent loading when agent files are unparsable
871
872## 1.0.80
873
874- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
875
876## 1.0.77
877
878- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
879- SDK: Add session support and permission denial tracking
880- Fix token limit errors in conversation summarization
881- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
882
883## 1.0.73
884
885- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
886- MCP: Press Esc to cancel OAuth authentication flows
887- Bash: Improved command validation and reduced false security warnings
888- UI: Enhanced spinner animations and status line visual hierarchy
889- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
890
891## 1.0.72
892
893- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
894
895## 1.0.71
896
897- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
898- Customizable status line: add your terminal prompt to Claude Code with /statusline
899
900## 1.0.70
901
902- Performance: Optimized message rendering for better performance with large contexts
903- Windows: Fixed native file search, ripgrep, and subagent functionality
904- Added support for @-mentions in slash command arguments
905
906## 1.0.69
907
908- Upgraded Opus to version 4.1
909
910## 1.0.68
911
912- Fix incorrect model names being used for certain commands like `/pr-comments`
913- 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.
914- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
915- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
916- SDK: Added canUseTool callback support for tool confirmation
917- Added `disableAllHooks` setting
918- Improved file suggestions performance in large repos
919
920## 1.0.65
921
922- IDE: Fixed connection stability issues and error handling for diagnostics
923- Windows: Fixed shell environment setup for users without .bashrc files
924
925## 1.0.64
926
927- Agents: Added model customization support - you can now specify which model an agent should use
928- Agents: Fixed unintended access to the recursive agent tool
929- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
930- SDK: Fixed user input tracking across multi-turn conversations
931- Added hidden files to file search and @-mention suggestions
932
933## 1.0.63
934
935- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
936
937## 1.0.62
938
939- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
940- Hooks: Added SessionStart hook for new session initialization
941- /add-dir command now supports typeahead for directory paths
942- Improved network connectivity check reliability
943
944## 1.0.61
945
946- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
947- Settings: Added `--settings` flag to load settings from a JSON file
948- Settings: Fixed resolution of settings files paths that are symlinks
949- OTEL: Fixed reporting of wrong organization after authentication changes
950- Slash commands: Fixed permissions checking for allowed-tools with Bash
951- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
952- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
953- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
954
955## 1.0.60
956
957- You can now create custom subagents for specialized tasks! Run /agents to get started
958
959## 1.0.59
960
961- SDK: Added tool confirmation support with canUseTool callback
962- SDK: Allow specifying env for spawned process
963- Hooks: Exposed PermissionDecision to hooks (including "ask")
964- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
965- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
966
967## 1.0.58
968
969- Added support for reading PDFs
970- MCP: Improved server health status display in 'claude mcp list'
971- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
972
973## 1.0.57
974
975- Added support for specifying a model in slash commands
976- Improved permission messages to help Claude understand allowed tools
977- Fix: Remove trailing newlines from bash output in terminal wrapping
978
979## 1.0.56
980
981- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
982- Fixes for WSL IDE detection
983- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
984
985## 1.0.55
986
987- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
988- Windows: fixed Ctrl+Z crash
989- SDK: Added ability to capture error logging
990- Add --system-prompt-file option to override system prompt in print mode
991
992## 1.0.54
993
994- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
995- Custom slash commands: Added argument-hint to frontmatter
996- Windows: OAuth uses port 45454 and properly constructs browser URL
997- Windows: mode switching now uses alt + m, and plan mode renders properly
998- Shell: Switch to in-memory shell snapshot to fix file-related errors
999
1000## 1.0.53
1001
1002- Updated @-mention file truncation from 100 lines to 2000 lines
1003- 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).
1004
1005## 1.0.52
1006
1007- Added support for MCP server instructions
1008
1009## 1.0.51
1010
1011- Added support for native Windows (requires Git for Windows)
1012- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
1013- Settings: /doctor can now help you identify and fix invalid setting files
1014- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
1015- Increased auto-compact warning threshold from 60% to 80%
1016- Fixed an issue with handling user directories with spaces for shell snapshots
1017- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
1018- Custom slash commands: Fixed user-level commands in subdirectories
1019- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
1020
1021## 1.0.48
1022
1023- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
1024- Added progress messages to Bash tool based on the last 5 lines of command output
1025- Added expanding variables support for MCP server configuration
1026- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
1027- Improved IDE extension path handling when Claude Code runs in WSL
1028- Hooks: Added a PreCompact hook
1029- Vim mode: Added c, f/F, t/T
1030
1031## 1.0.45
1032
1033- Redesigned Search (Grep) tool with new tool input parameters and features
1034- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
1035- Fixed config file corruption issue by enforcing atomic writes
1036- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
1037- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
1038- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
1039
1040## 1.0.44
1041
1042- New /export command lets you quickly export a conversation for sharing
1043- MCP: resource_link tool results are now supported
1044- MCP: tool annotations and tool titles now display in /mcp view
1045- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
1046
1047## 1.0.43
1048
1049- Fixed a bug where the theme selector was saving excessively
1050- Hooks: Added EPIPE system error handling
1051
1052## 1.0.42
1053
1054- Added tilde (`~`) expansion support to `/add-dir` command
1055
1056## 1.0.41
1057
1058- Hooks: Split Stop hook triggering into Stop and SubagentStop
1059- Hooks: Enabled optional timeout configuration for each command
1060- Hooks: Added "hook_event_name" to hook input
1061- Fixed a bug where MCP tools would display twice in tool list
1062- New tool parameters JSON for Bash tool in `tool_decision` event
1063
1064## 1.0.40
1065
1066- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1067
1068## 1.0.39
1069
1070- New Active Time metric in OpenTelemetry logging
1071
1072## 1.0.38
1073
1074- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1075
1076## 1.0.37
1077
1078- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1079
1080## 1.0.36
1081
1082- Web search now takes today's date into context
1083- Fixed a bug where stdio MCP servers were not terminating properly on exit
1084
1085## 1.0.35
1086
1087- Added support for MCP OAuth Authorization Server discovery
1088
1089## 1.0.34
1090
1091- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1092
1093## 1.0.33
1094
1095- Improved logging functionality with session ID support
1096- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1097- Improvements to plan mode
1098
1099## 1.0.32
1100
1101- Updated loopback config for litellm
1102- Added forceLoginMethod setting to bypass login selection screen
1103
1104## 1.0.31
1105
1106- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1107
1108## 1.0.30
1109
1110- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1111- Improved file path autocomplete with filename matching
1112- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1113- Enhanced jq regex support for complex filters with pipes and select
1114
1115## 1.0.29
1116
1117- Improved CJK character support in cursor navigation and rendering
1118
1119## 1.0.28
1120
1121- Slash commands: Fix selector display during history navigation
1122- Resizes images before upload to prevent API size limit errors
1123- Added XDG_CONFIG_HOME support to configuration directory
1124- Performance optimizations for memory usage
1125- New attributes (terminal.type, language) in OpenTelemetry logging
1126
1127## 1.0.27
1128
1129- Streamable HTTP MCP servers are now supported
1130- Remote MCP servers (SSE and HTTP) now support OAuth
1131- MCP resources can now be @-mentioned
1132- /resume slash command to switch conversations within Claude Code
1133
1134## 1.0.25
1135
1136- Slash commands: moved "project" and "user" prefixes to descriptions
1137- Slash commands: improved reliability for command discovery
1138- Improved support for Ghostty
1139- Improved web search reliability
1140
1141## 1.0.24
1142
1143- Improved /mcp output
1144- Fixed a bug where settings arrays got overwritten instead of merged
1145
1146## 1.0.23
1147
1148- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1149- Released Python SDK: pip install claude-code-sdk to get started
1150
1151## 1.0.22
1152
1153- SDK: Renamed `total_cost` to `total_cost_usd`
1154
1155## 1.0.21
1156
1157- Improved editing of files with tab-based indentation
1158- Fix for tool_use without matching tool_result errors
1159- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1160
1161## 1.0.18
1162
1163- Added --add-dir CLI argument for specifying additional working directories
1164- Added streaming input support without require -p flag
1165- Improved startup performance and session storage performance
1166- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1167- Added detailed MCP server tools display (/mcp)
1168- MCP authentication and permission improvements
1169- Added auto-reconnection for MCP SSE connections on disconnect
1170- Fixed issue where pasted content was lost when dialogs appeared
1171
1172## 1.0.17
1173
1174- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1175- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1176- MCP server list UI improvements
1177- Update Claude Code process title to display "claude" instead of "node"
1178
1179## 1.0.11
1180
1181- Claude Code can now also be used with a Claude Pro subscription
1182- Added /upgrade for smoother switching to Claude Max plans
1183- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1184- Improved shell configuration error handling
1185- Improved todo list handling during compaction
1186
1187## 1.0.10
1188
1189- Added markdown table support
1190- Improved streaming performance
1191
1192## 1.0.8
1193
1194- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1195- Increased default otel interval from 1s -> 5s
1196- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1197- Fixed a regression where search tools unnecessarily asked for permissions
1198- Added support for triggering thinking non-English languages
1199- Improved compacting UI
1200
1201## 1.0.7
1202
1203- Renamed /allowed-tools -> /permissions
1204- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1205- Deprecated claude config commands in favor of editing settings.json
1206- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1207- Improved error handling for /install-github-app
1208- Bugfixes, UI polish, and tool reliability improvements
1209
1210## 1.0.6
1211
1212- Improved edit reliability for tab-indented files
1213- Respect CLAUDE_CONFIG_DIR everywhere
1214- Reduced unnecessary tool permission prompts
1215- Added support for symlinks in @file typeahead
1216- Bugfixes, UI polish, and tool reliability improvements
1217
1218## 1.0.4
1219
1220- Fixed a bug where MCP tool errors weren't being parsed correctly
1221
1222## 1.0.1
1223
1224- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1225- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1226- Updated documentation links and OAuth process descriptions
1227
1228## 1.0.0
1229
1230- Claude Code is now generally available
1231- Introducing Sonnet 4 and Opus 4 models
1232
1233## 0.2.125
1234
1235- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1236- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1237
1238## 0.2.117
1239
1240- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1241- Introduced settings.cleanupPeriodDays
1242- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1243- Introduced --debug mode
1244
1245## 0.2.108
1246
1247- You can now send messages to Claude while it works to steer Claude in real-time
1248- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1249- Fixed a bug where thinking was not working in -p mode
1250- Fixed a regression in /cost reporting
1251- Deprecated MCP wizard interface in favor of other MCP commands
1252- Lots of other bugfixes and improvements
1253
1254## 0.2.107
1255
1256- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1257
1258## 0.2.106
1259
1260- MCP SSE server configs can now specify custom headers
1261- Fixed a bug where MCP permission prompt didn't always show correctly
1262
1263## 0.2.105
1264
1265- Claude can now search the web
1266- Moved system & account status to /status
1267- Added word movement keybindings for Vim
1268- Improved latency for startup, todo tool, and file edits
1269
1270## 0.2.102
1271
1272- Improved thinking triggering reliability
1273- Improved @mention reliability for images and folders
1274- You can now paste multiple large chunks into one prompt
1275
1276## 0.2.100
1277
1278- Fixed a crash caused by a stack overflow error
1279- Made db storage optional; missing db support disables --continue and --resume
1280
1281## 0.2.98
1282
1283- Fixed an issue where auto-compact was running twice
1284
1285## 0.2.96
1286
1287- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1288
1289## 0.2.93
1290
1291- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1292- Claude now has access to a Todo list that helps it stay on track and be more organized
1293
1294## 0.2.82
1295
1296- Added support for --disallowedTools
1297- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1298
1299## 0.2.75
1300
1301- Hit Enter to queue up additional messages while Claude is working
1302- Drag in or copy/paste image files directly into the prompt
1303- @-mention files to directly add them to context
1304- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1305- Improved performance for filename auto-complete
1306
1307## 0.2.74
1308
1309- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1310- Task tool can now perform writes and run bash commands
1311
1312## 0.2.72
1313
1314- Updated spinner to indicate tokens loaded and tool usage
1315
1316## 0.2.70
1317
1318- Network commands like curl are now available for Claude to use
1319- Claude can now run multiple web queries in parallel
1320- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1321
1322## 0.2.69
1323
1324- Fixed UI glitches with improved Select component behavior
1325- Enhanced terminal output display with better text truncation logic
1326
1327## 0.2.67
1328
1329- Shared project permission rules can be saved in .claude/settings.json
1330
1331## 0.2.66
1332
1333- Print mode (-p) now supports streaming output via --output-format=stream-json
1334- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1335
1336## 0.2.63
1337
1338- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1339
1340## 0.2.61
1341
1342- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1343- Enhanced image detection for more reliable clipboard paste functionality
1344- Fixed an issue where ESC key could crash the conversation history selector
1345
1346## 0.2.59
1347
1348- Copy+paste images directly into your prompt
1349- Improved progress indicators for bash and fetch tools
1350- Bugfixes for non-interactive mode (-p)
1351
1352## 0.2.54
1353
1354- Quickly add to Memory by starting your message with '#'
1355- Press ctrl+r to see full output for long tool results
1356- Added support for MCP SSE transport
1357
1358## 0.2.53
1359
1360- New web fetch tool lets Claude view URLs that you paste in
1361- Fixed a bug with JPEG detection
1362
1363## 0.2.50
1364
1365- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1366
1367## 0.2.49
1368
1369- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1370
1371## 0.2.47
1372
1373- Press Tab to auto-complete file and folder names
1374- Press Shift + Tab to toggle auto-accept for file edits
1375- Automatic conversation compaction for infinite conversation length (toggle with /config)
1376
1377## 0.2.44
1378
1379- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1380
1381## 0.2.41
1382
1383- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1384- MCP server startup no longer blocks the app from starting up
1385
1386## 0.2.37
1387
1388- New /release-notes command lets you view release notes at any time
1389- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1390
1391## 0.2.36
1392
1393- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1394- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1395
1396## 0.2.34
1397
1398- Vim bindings for text input - enable with /vim or /config
1399
1400## 0.2.32
1401
1402- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1403- Fix for some PersistentShell issues
1404
1405## 0.2.31
1406
1407- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1408- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1409
1410## 0.2.30
1411
1412- Added ANSI color theme for better terminal compatibility
1413- Fixed issue where slash command arguments weren't being sent properly
1414- (Mac-only) API keys are now stored in macOS Keychain
1415
1416## 0.2.26
1417
1418- New /approved-tools command for managing tool permissions
1419- Word-level diff display for improved code readability
1420- Fuzzy matching for slash commands
1421
1422## 0.2.21
1423
1424- Fuzzy matching for /commands