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 5aac2b1 · on Jan 7, 2026 History
1,298 lines · 60.0 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.0
4
5- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
6- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
7- Added support for `agent` field in skills to specify agent type for execution
8- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
9- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
10- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
11- Added `CLAUDE_CODE_HIDE_ACCOUNT_INFO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
12- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
13- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
14- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
15- Fixed Esc key with queued prompts to only move them to input without canceling the running task
16- Reduced permission prompts for complex bash commands
17- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
18- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
19- Fixed plugin path resolution for file-based marketplace sources
20- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
21- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
22- Fixed Claude in Chrome support for WSL environments
23- Fixed Windows native installer silently failing when executable creation fails
24- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
25- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
26- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
27- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
28- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
29- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
30- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
31- Added hooks support for skill and slash command frontmatter
32- 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
33- Added `/plan` command shortcut to enable plan mode directly from the prompt
34- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
35- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
36- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
37- Added support for `once: true` config for hooks
38- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
39- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
40- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
41- Added left/right arrow key navigation for cycling through tabs in dialogs
42- Added real-time thinking block display in Ctrl+O transcript mode
43- Added filepath to full output in background bash task details dialog
44- Added Skills as a separate category in the context visualization
45- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
46- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
47- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
48- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
49- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
50- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
51- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
52- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
53- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
54- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
55- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
56- Fixed commands with `$()` command substitution failing with parse errors
57- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
58- 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)
59- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
60- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
61- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
62- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
63- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
64- Fixed subagents sometimes not inheriting the parent's model by default
65- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
66- Fixed duplicate Bash commands appearing in permission request option labels
67- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
68- Fixed background task completion notifications to appear proactively with bullet point
69- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
70- Fixed cursor disappearing after dismissing permission dialogs
71- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
72- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
73- Fixed images being silently dropped when queueing messages while backgrounding a task
74- Fixed large pasted images failing with "Image was too large" error
75- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
76- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
77- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
78- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
79- Fixed spinner token counter to properly accumulate tokens from subagents during execution
80- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
81- Fixed race condition where LSP tool could return "no server available" during startup
82- Fixed feedback submission hanging indefinitely when network requests timeout
83- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
84- Fixed hook success message showing trailing colon when hook has no output
85- Multiple optimizations to improve startup performance
86- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
87- Improved performance when reading Jupyter notebooks with many cells
88- Improved reliability for piped input like `cat refactor.md | claude`
89- Improved reliability for AskQuestion tool
90- Improved sed in-place edit commands to render as file edits with diff preview
91- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
92- Improved compaction reliability
93- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
94- Improved skills to show progress while executing, displaying tool uses as they happen
95- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
96- Improved skill suggestions to prioritize recently and frequently used skills
97- Improved spinner feedback when waiting for the first response token
98- Improved token count display in spinner to include tokens from background agents
99- Improved incremental output for async agents to give the main thread more control and visibility
100- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
101- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
102- Improved macOS screenshot paste reliability with TIFF format support
103- Improved `/stats` output
104- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
105- Changed "Interrupted" message color from red to grey for a less alarming appearance
106- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
107- Removed underline styling from image reference links
108- [SDK] Changed minimum zod peer dependency to ^4.0.0
109- [VSCode] Added currently selected model name to the context menu
110- [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")
111- [VSCode] Fixed paragraph breaks not rendering in markdown content
112- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
113- [Windows] Fixed issue with improper rendering
114
115## 2.0.76
116
117- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
118
119## 2.0.75
120
121- Minor bugfixes
122
123## 2.0.74
124
125- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
126- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
127- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
128- Added syntax highlighting info to theme picker
129- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
130- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
131- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
132- Fixed a potential crash when syntax highlighting isn't initialized correctly
133- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
134- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
135- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
136- [Windows] Fixed issue with improper rendering
137- [VSCode] Added gift tag pictogram for year-end promotion message
138
139## 2.0.73
140
141- Added clickable `[Image #N]` links that open attached images in the default viewer
142- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
143- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
144- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
145- Fixed slow input history cycling and race condition that could overwrite text after message submission
146- Improved `/theme` command to open theme picker directly
147- Improved theme picker UI
148- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
149- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
150
151## 2.0.72
152
153- 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
154- Reduced terminal flickering
155- Added scannable QR code to mobile app tip for quick app downloads
156- Added loading indicator when resuming conversations for better feedback
157- Fixed `/context` command not respecting custom system prompts in non-interactive mode
158- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
159- Improved @ mention file suggestion speed (~3x faster in git repositories)
160- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
161- Improved settings validation errors to be more prominent
162- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
163
164## 2.0.71
165
166- Added /config toggle to enable/disable prompt suggestions
167- Added `/settings` as an alias for the `/config` command
168- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
169- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
170- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
171- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
172- New syntax highlighting engine for native build
173
174## 2.0.70
175
176- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
177- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
178- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
179- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
180- Fixed input being cleared when processing queued commands while the user was typing
181- Fixed prompt suggestions replacing typed input when pressing Tab
182- Fixed diff view not updating when terminal is resized
183- Improved memory usage by 3x for large conversations
184- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
185- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
186- Fix thinking mode toggle in /config not persisting correctly
187- Improve UI for file creation permission dialog
188
189## 2.0.69
190
191- Minor bugfixes
192
193## 2.0.68
194
195- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
196- Fixed a bug where disallowed MCP tools were visible to the model
197- Fixed an issue where steering messages could be lost while a subagent is working
198- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
199- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
200- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
201
202## 2.0.67
203
204- Thinking mode is now enabled by default for Opus 4.5
205- Thinking mode configuration has moved to /config
206- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
207- Show reason why autoupdater is disabled in `/doctor`
208- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
209- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
210- Fixed scroll position resetting after deleting a permission rule in `/permissions`
211- 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
212- Fixed `claude install --force` not bypassing stale lock files
213- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
214- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
215
216## 2.0.65
217
218- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
219- Added context window information to status line input
220- Added `fileSuggestion` setting for custom `@` file search commands
221- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
222- Fixed prompt not being saved to history when aborting a query with Escape
223- Fixed Read tool image handling to identify format from bytes instead of file extension
224
225## 2.0.64
226
227- Made auto-compacting instant
228- Agents and bash commands can run asynchronously and send messages to wake up the main agent
229- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
230- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
231- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
232- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
233- Fixed auto-loading .env when using native installer
234- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
235- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
236- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
237- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
238- Bedrock: Improve efficiency of token counting
239- Bedrock: Add support for `aws login` AWS Management Console credentials
240- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
241
242## 2.0.62
243
244- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
245- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
246- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
247- Fixed slash command selection not working when multiple commands share the same name
248- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
249- Fixed running versions getting removed because lock file incorrectly going stale
250- Fixed IDE diff tab not closing when rejecting file changes
251
252## 2.0.61
253
254- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
255
256## 2.0.60
257
258- Added background agent support. Agents run in the background while you work
259- Added --disable-slash-commands CLI flag to disable all slash commands
260- Added model name to "Co-Authored-By" commit messages
261- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
262- Updated Fetch to skip summarization for pre-approved websites
263- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
264
265## 2.0.59
266
267- Added --agent CLI flag to override the agent setting for the current session
268- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
269- VS Code: Fixed .claude.json config file being read from incorrect location
270
271## 2.0.58
272
273- Pro users now have access to Opus 4.5 as part of their subscription!
274- Fixed timer duration showing "11m 60s" instead of "12m 0s"
275- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
276
277## 2.0.57
278
279- Added feedback input when rejecting plans, allowing users to tell Claude what to change
280- VSCode: Added streaming message support for real-time response display
281
282## 2.0.56
283
284- Added setting to enable/disable terminal progress bar (OSC 9;4)
285- 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.
286
287## 2.0.55
288
289- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
290- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
291- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
292- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
293
294## 2.0.54
295
296- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
297- Fix issue with excessive iTerm notifications
298
299## 2.0.52
300
301- Fixed duplicate message display when starting Claude with a command line argument
302- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
303- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
304- Permit some uses of `$!` in bash commands
305
306## 2.0.51
307
308- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
309- Introducing Claude Code for Desktop: https://claude.com/download
310- 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
311- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
312- Plan Mode now builds more precise plans and executes more thoroughly
313- Usage limit notifications now easier to understand
314- Switched `/usage` back to "% used"
315- Fixed handling of thinking errors
316- Fixed performance regression
317
318## 2.0.50
319
320- Fixed bug preventing calling MCP tools that have nested references in their input schemas
321- Silenced a noisy but harmless error during upgrades
322- Improved ultrathink text display
323- Improved clarity of 5-hour session limit warning message
324
325## 2.0.49
326
327- Added readline-style ctrl-y for pasting deleted text
328- Improved clarity of usage limit warning message
329- Fixed handling of subagent permissions
330
331## 2.0.47
332
333- Improved error messages and validation for `claude --teleport`
334- Improved error handling in `/usage`
335- Fixed race condition with history entry not getting logged at exit
336- Fixed Vertex AI configuration not being applied from `settings.json`
337
338## 2.0.46
339
340- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
341
342## 2.0.45
343
344- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
345- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
346- Send background tasks to Claude Code on the web by starting a message with `&`
347
348## 2.0.43
349
350- Added `permissionMode` field for custom agents
351- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
352- Added skills frontmatter field to declare skills to auto-load for subagents
353- Added the `SubagentStart` hook event
354- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
355- Fixed duplicate rendering of some messages in the UI
356- Fixed some visual flickers
357- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
358
359## 2.0.42
360
361- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
362
363## 2.0.41
364
365- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
366- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
367- Fixed incorrect labeling of user settings vs project settings in command descriptions
368- Fixed crash when plugin command hooks timeout during execution
369- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
370- Fixed broken security documentation links in trust dialogs and onboarding
371- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
372- ctrl-r history search landing on a slash command no longer cancels the search
373- SDK: Support custom timeouts for hooks
374- Allow more safe git commands to run without approval
375- Plugins: Added support for sharing and installing output styles
376- Teleporting a session from web will automatically set the upstream branch
377
378## 2.0.37
379
380- Fixed how idleness is computed for notifications
381- Hooks: Added matcher values for Notification hook events
382- Output Styles: Added `keep-coding-instructions` option to frontmatter
383
384## 2.0.36
385
386- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
387- Fixed queued messages being incorrectly executed as bash commands
388- Fixed input being lost when typing while a queued message is processed
389
390## 2.0.35
391
392- Improve fuzzy search results when searching commands
393- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
394- 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
395- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
396- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
397
398## 2.0.34
399
400- VSCode Extension: Added setting to configure the initial permission mode for new conversations
401- Improved file path suggestion performance with native Rust-based fuzzy finder
402- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
403- Fixed memory crash when reading or writing large files (especially base64-encoded images)
404
405## 2.0.33
406
407- Native binary installs now launch quicker.
408- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
409- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
410
411## 2.0.32
412
413- Un-deprecate output styles based on community feedback
414- Added `companyAnnouncements` setting for displaying announcements on startup
415- Fixed hook progress messages not updating correctly during PostToolUse hook execution
416
417## 2.0.31
418
419- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
420- Vertex: add support for Web Search on supported models
421- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
422- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
423- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
424- Fixed plugin uninstall not removing plugins
425
426## 2.0.30
427
428- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
429- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
430- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
431- Added prompt-based stop hooks
432- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
433- Enabled SSE MCP servers on native build
434- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
435- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
436- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
437- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
438- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
439- Fixed bug that saved session permissions to local settings
440- Fixed MCP tools not being available to sub-agents
441- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
442- Fixed delay when navigating through typeahead suggestions with arrow keys
443- VSCode: Restored selection indicator in input footer showing current file or code selection status
444
445## 2.0.28
446
447- Plan mode: introduced new Plan subagent
448- Subagents: claude can now choose to resume subagents
449- Subagents: claude can dynamically choose the model used by its subagents
450- SDK: added --max-budget-usd flag
451- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
452- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
453- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
454- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
455- Various other bug fixes
456
457## 2.0.27
458
459- New UI for permission prompts
460- Added current branch filtering and search to session resume screen for easier navigation
461- Fixed directory @-mention causing "No assistant message found" error
462- VSCode Extension: Add config setting to include .gitignored files in file searches
463- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
464
465## 2.0.25
466
467- 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
468
469## 2.0.24
470
471- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
472- Claude Code Web: Support for Web -> CLI teleport
473- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
474- Bedrock: Display awsAuthRefresh output when auth is required
475
476## 2.0.22
477
478- Fixed content layout shift when scrolling through slash commands
479- IDE: Add toggle to enable/disable thinking.
480- Fix bug causing duplicate permission prompts with parallel tool calls
481- Add support for enterprise managed MCP allowlist and denylist
482
483## 2.0.21
484
485- Support MCP `structuredContent` field in tool responses
486- Added an interactive question tool
487- Claude will now ask you questions more often in plan mode
488- Added Haiku 4.5 as a model option for Pro users
489- Fixed an issue where queued commands don't have access to previous messages' output
490
491## 2.0.20
492
493- Added support for Claude Skills
494
495## 2.0.19
496
497- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
498- Fixed a bug where Haiku was unnecessarily called in print mode
499
500## 2.0.17
501
502- Added Haiku 4.5 to model selector!
503- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
504- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
505- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
506- OTEL: support HTTP_PROXY and HTTPS_PROXY
507- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
508
509## 2.0.15
510
511- Fixed bug with resuming where previously created files needed to be read again before writing
512- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
513
514## 2.0.14
515
516- Fix @-mentioning MCP servers to toggle them on/off
517- Improve permission checks for bash with inline env vars
518- Fix ultrathink + thinking toggle
519- Reduce unnecessary logins
520- Document --system-prompt
521- Several improvements to rendering
522- Plugins UI polish
523
524## 2.0.13
525
526- Fixed `/plugin` not working on native build
527
528## 2.0.12
529
530- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
531- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
532- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
533- `/plugin validate` command for validating plugin structure and configuration
534- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
535- Plugin documentation available at https://code.claude.com/docs/en/plugins
536- Comprehensive error messages and diagnostics via `/doctor` command
537- Avoid flickering in `/model` selector
538- Improvements to `/help`
539- Avoid mentioning hooks in `/resume` summaries
540- Changes to the "verbose" setting in `/config` now persist across sessions
541
542## 2.0.11
543
544- Reduced system prompt size by 1.4k tokens
545- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
546- Fixed Opus fallback rate limit errors appearing incorrectly
547- Fixed /add-dir command selecting wrong default tab
548
549## 2.0.10
550
551- Rewrote terminal renderer for buttery smooth UI
552- Enable/disable MCP servers by @mentioning, or in /mcp
553- Added tab completion for shell commands in bash mode
554- PreToolUse hooks can now modify tool inputs
555- Press Ctrl-G to edit your prompt in your system's configured text editor
556- Fixes for bash permission checks with environment variables in the command
557
558## 2.0.9
559
560- Fix regression where bash backgrounding stopped working
561
562## 2.0.8
563
564- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
565- IDE: Add drag-and-drop support for files and folders in chat
566- /context: Fix counting for thinking blocks
567- Improve message rendering for users with light themes on dark terminals
568- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
569
570## 2.0.5
571
572- IDE: Fix IME unintended message submission with Enter and Tab
573- IDE: Add "Open in Terminal" link in login screen
574- Fix unhandled OAuth expiration 401 API errors
575- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
576
577## 2.0.1
578
579- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
580- Various bug fixes and presentation improvements
581
582## 2.0.0
583
584- New native VS Code extension
585- Fresh coat of paint throughout the whole app
586- /rewind a conversation to undo code changes
587- /usage command to see plan limits
588- Tab to toggle thinking (sticky across sessions)
589- Ctrl-R to search history
590- Unshipped claude config command
591- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
592- SDK: The Claude Code SDK is now the Claude Agent SDK
593- Add subagents dynamically with `--agents` flag
594
595## 1.0.126
596
597- Enable /context command for Bedrock and Vertex
598- Add mTLS support for HTTP-based OpenTelemetry exporters
599
600## 1.0.124
601
602- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
603- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
604- No longer inform Claude of the list of allowed tools when permission is denied
605- Fixed security vulnerability in Bash tool permission checks
606- Improved VSCode extension performance for large files
607
608## 1.0.123
609
610- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
611- Fixed thinking mode triggering on negation phrases like "don't think"
612- Fixed rendering performance degradation during token streaming
613- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
614- Enhanced BashTool environment snapshot logging
615- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
616- Migrated --debug logging to a file, to enable easy tailing & filtering
617
618## 1.0.120
619
620- Fix input lag during typing, especially noticeable with large prompts
621- Improved VSCode extension command registry and sessions dialog user experience
622- Enhanced sessions dialog responsiveness and visual feedback
623- Fixed IDE compatibility issue by removing worktree support check
624- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
625
626## 1.0.119
627
628- Fix Windows issue where process visually freezes on entering interactive mode
629- Support dynamic headers for MCP servers via headersHelper configuration
630- Fix thinking mode not working in headless sessions
631- Fix slash commands now properly update allowed tools instead of replacing them
632
633## 1.0.117
634
635- Add Ctrl-R history search to recall previous commands like bash/zsh
636- Fix input lag while typing, especially on Windows
637- Add sed command to auto-allowed commands in acceptEdits mode
638- Fix Windows PATH comparison to be case-insensitive for drive letters
639- Add permissions management hint to /add-dir output
640
641## 1.0.115
642
643- Improve thinking mode display with enhanced visual effects
644- Type /t to temporarily disable thinking mode in your prompt
645- Improve path validation for glob and grep tools
646- Show condensed output for post-tool hooks to reduce visual clutter
647- Fix visual feedback when loading state completes
648- Improve UI consistency for permission request dialogs
649
650## 1.0.113
651
652- Deprecated piped input in interactive mode
653- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
654
655## 1.0.112
656
657- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
658- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
659- Hooks: Added systemMessage support for SessionEnd hooks
660- Added `spinnerTipsEnabled` setting to disable spinner tips
661- IDE: Various improvements and bug fixes
662
663## 1.0.111
664
665- /model now validates provided model names
666- Fixed Bash tool crashes caused by malformed shell syntax parsing
667
668## 1.0.110
669
670- /terminal-setup command now supports WezTerm
671- MCP: OAuth tokens now proactively refresh before expiration
672- Fixed reliability issues with background Bash processes
673
674## 1.0.109
675
676- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
677
678## 1.0.106
679
680- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
681
682## 1.0.97
683
684- Settings: /doctor now validates permission rule syntax and suggests corrections
685
686## 1.0.94
687
688- Vertex: add support for global endpoints for supported models
689- /memory command now allows direct editing of all imported memory files
690- SDK: Add custom tools as callbacks
691- Added /todos command to list current todo items
692
693## 1.0.93
694
695- Windows: Add alt + v shortcut for pasting images from clipboard
696- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
697
698## 1.0.90
699
700- Settings file changes take effect immediately - no restart required
701
702## 1.0.88
703
704- Fixed issue causing "OAuth authentication is currently not supported"
705- Status line input now includes `exceeds_200k_tokens`
706- Fixed incorrect usage tracking in /cost.
707- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
708- Bedrock: Updated default Sonnet model to Sonnet 4
709
710## 1.0.86
711
712- Added /context to help users self-serve debug context issues
713- SDK: Added UUID support for all SDK messages
714- SDK: Added `--replay-user-messages` to replay user messages back to stdout
715
716## 1.0.85
717
718- Status line input now includes session cost info
719- Hooks: Introduced SessionEnd hook
720
721## 1.0.84
722
723- Fix tool_use/tool_result id mismatch error when network is unstable
724- Fix Claude sometimes ignoring real-time steering when wrapping up a task
725- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
726- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
727
728## 1.0.83
729
730- @-mention: Support files with spaces in path
731- New shimmering spinner
732
733## 1.0.82
734
735- SDK: Add request cancellation support
736- SDK: New additionalDirectories option to search custom paths, improved slash command processing
737- Settings: Validation prevents invalid fields in .claude/settings.json files
738- MCP: Improve tool name consistency
739- Bash: Fix crash when Claude tries to automatically read large files
740
741## 1.0.81
742
743- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
744- Agents: Fix custom agent loading when agent files are unparsable
745
746## 1.0.80
747
748- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
749
750## 1.0.77
751
752- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
753- SDK: Add session support and permission denial tracking
754- Fix token limit errors in conversation summarization
755- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
756
757## 1.0.73
758
759- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
760- MCP: Press Esc to cancel OAuth authentication flows
761- Bash: Improved command validation and reduced false security warnings
762- UI: Enhanced spinner animations and status line visual hierarchy
763- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
764
765## 1.0.72
766
767- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
768
769## 1.0.71
770
771- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
772- Customizable status line: add your terminal prompt to Claude Code with /statusline
773
774## 1.0.70
775
776- Performance: Optimized message rendering for better performance with large contexts
777- Windows: Fixed native file search, ripgrep, and subagent functionality
778- Added support for @-mentions in slash command arguments
779
780## 1.0.69
781
782- Upgraded Opus to version 4.1
783
784## 1.0.68
785
786- Fix incorrect model names being used for certain commands like `/pr-comments`
787- 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.
788- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
789- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
790- SDK: Added canUseTool callback support for tool confirmation
791- Added `disableAllHooks` setting
792- Improved file suggestions performance in large repos
793
794## 1.0.65
795
796- IDE: Fixed connection stability issues and error handling for diagnostics
797- Windows: Fixed shell environment setup for users without .bashrc files
798
799## 1.0.64
800
801- Agents: Added model customization support - you can now specify which model an agent should use
802- Agents: Fixed unintended access to the recursive agent tool
803- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
804- SDK: Fixed user input tracking across multi-turn conversations
805- Added hidden files to file search and @-mention suggestions
806
807## 1.0.63
808
809- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
810
811## 1.0.62
812
813- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
814- Hooks: Added SessionStart hook for new session initialization
815- /add-dir command now supports typeahead for directory paths
816- Improved network connectivity check reliability
817
818## 1.0.61
819
820- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
821- Settings: Added `--settings` flag to load settings from a JSON file
822- Settings: Fixed resolution of settings files paths that are symlinks
823- OTEL: Fixed reporting of wrong organization after authentication changes
824- Slash commands: Fixed permissions checking for allowed-tools with Bash
825- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
826- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
827- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
828
829## 1.0.60
830
831- You can now create custom subagents for specialized tasks! Run /agents to get started
832
833## 1.0.59
834
835- SDK: Added tool confirmation support with canUseTool callback
836- SDK: Allow specifying env for spawned process
837- Hooks: Exposed PermissionDecision to hooks (including "ask")
838- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
839- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
840
841## 1.0.58
842
843- Added support for reading PDFs
844- MCP: Improved server health status display in 'claude mcp list'
845- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
846
847## 1.0.57
848
849- Added support for specifying a model in slash commands
850- Improved permission messages to help Claude understand allowed tools
851- Fix: Remove trailing newlines from bash output in terminal wrapping
852
853## 1.0.56
854
855- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
856- Fixes for WSL IDE detection
857- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
858
859## 1.0.55
860
861- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
862- Windows: fixed Ctrl+Z crash
863- SDK: Added ability to capture error logging
864- Add --system-prompt-file option to override system prompt in print mode
865
866## 1.0.54
867
868- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
869- Custom slash commands: Added argument-hint to frontmatter
870- Windows: OAuth uses port 45454 and properly constructs browser URL
871- Windows: mode switching now uses alt + m, and plan mode renders properly
872- Shell: Switch to in-memory shell snapshot to fix file-related errors
873
874## 1.0.53
875
876- Updated @-mention file truncation from 100 lines to 2000 lines
877- 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).
878
879## 1.0.52
880
881- Added support for MCP server instructions
882
883## 1.0.51
884
885- Added support for native Windows (requires Git for Windows)
886- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
887- Settings: /doctor can now help you identify and fix invalid setting files
888- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
889- Increased auto-compact warning threshold from 60% to 80%
890- Fixed an issue with handling user directories with spaces for shell snapshots
891- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
892- Custom slash commands: Fixed user-level commands in subdirectories
893- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
894
895## 1.0.48
896
897- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
898- Added progress messages to Bash tool based on the last 5 lines of command output
899- Added expanding variables support for MCP server configuration
900- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
901- Improved IDE extension path handling when Claude Code runs in WSL
902- Hooks: Added a PreCompact hook
903- Vim mode: Added c, f/F, t/T
904
905## 1.0.45
906
907- Redesigned Search (Grep) tool with new tool input parameters and features
908- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
909- Fixed config file corruption issue by enforcing atomic writes
910- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
911- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
912- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
913
914## 1.0.44
915
916- New /export command lets you quickly export a conversation for sharing
917- MCP: resource_link tool results are now supported
918- MCP: tool annotations and tool titles now display in /mcp view
919- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
920
921## 1.0.43
922
923- Fixed a bug where the theme selector was saving excessively
924- Hooks: Added EPIPE system error handling
925
926## 1.0.42
927
928- Added tilde (`~`) expansion support to `/add-dir` command
929
930## 1.0.41
931
932- Hooks: Split Stop hook triggering into Stop and SubagentStop
933- Hooks: Enabled optional timeout configuration for each command
934- Hooks: Added "hook_event_name" to hook input
935- Fixed a bug where MCP tools would display twice in tool list
936- New tool parameters JSON for Bash tool in `tool_decision` event
937
938## 1.0.40
939
940- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
941
942## 1.0.39
943
944- New Active Time metric in OpenTelemetry logging
945
946## 1.0.38
947
948- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
949
950## 1.0.37
951
952- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
953
954## 1.0.36
955
956- Web search now takes today's date into context
957- Fixed a bug where stdio MCP servers were not terminating properly on exit
958
959## 1.0.35
960
961- Added support for MCP OAuth Authorization Server discovery
962
963## 1.0.34
964
965- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
966
967## 1.0.33
968
969- Improved logging functionality with session ID support
970- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
971- Improvements to plan mode
972
973## 1.0.32
974
975- Updated loopback config for litellm
976- Added forceLoginMethod setting to bypass login selection screen
977
978## 1.0.31
979
980- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
981
982## 1.0.30
983
984- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
985- Improved file path autocomplete with filename matching
986- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
987- Enhanced jq regex support for complex filters with pipes and select
988
989## 1.0.29
990
991- Improved CJK character support in cursor navigation and rendering
992
993## 1.0.28
994
995- Slash commands: Fix selector display during history navigation
996- Resizes images before upload to prevent API size limit errors
997- Added XDG_CONFIG_HOME support to configuration directory
998- Performance optimizations for memory usage
999- New attributes (terminal.type, language) in OpenTelemetry logging
1000
1001## 1.0.27
1002
1003- Streamable HTTP MCP servers are now supported
1004- Remote MCP servers (SSE and HTTP) now support OAuth
1005- MCP resources can now be @-mentioned
1006- /resume slash command to switch conversations within Claude Code
1007
1008## 1.0.25
1009
1010- Slash commands: moved "project" and "user" prefixes to descriptions
1011- Slash commands: improved reliability for command discovery
1012- Improved support for Ghostty
1013- Improved web search reliability
1014
1015## 1.0.24
1016
1017- Improved /mcp output
1018- Fixed a bug where settings arrays got overwritten instead of merged
1019
1020## 1.0.23
1021
1022- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1023- Released Python SDK: pip install claude-code-sdk to get started
1024
1025## 1.0.22
1026
1027- SDK: Renamed `total_cost` to `total_cost_usd`
1028
1029## 1.0.21
1030
1031- Improved editing of files with tab-based indentation
1032- Fix for tool_use without matching tool_result errors
1033- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1034
1035## 1.0.18
1036
1037- Added --add-dir CLI argument for specifying additional working directories
1038- Added streaming input support without require -p flag
1039- Improved startup performance and session storage performance
1040- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1041- Added detailed MCP server tools display (/mcp)
1042- MCP authentication and permission improvements
1043- Added auto-reconnection for MCP SSE connections on disconnect
1044- Fixed issue where pasted content was lost when dialogs appeared
1045
1046## 1.0.17
1047
1048- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1049- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1050- MCP server list UI improvements
1051- Update Claude Code process title to display "claude" instead of "node"
1052
1053## 1.0.11
1054
1055- Claude Code can now also be used with a Claude Pro subscription
1056- Added /upgrade for smoother switching to Claude Max plans
1057- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1058- Improved shell configuration error handling
1059- Improved todo list handling during compaction
1060
1061## 1.0.10
1062
1063- Added markdown table support
1064- Improved streaming performance
1065
1066## 1.0.8
1067
1068- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1069- Increased default otel interval from 1s -> 5s
1070- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1071- Fixed a regression where search tools unnecessarily asked for permissions
1072- Added support for triggering thinking non-English languages
1073- Improved compacting UI
1074
1075## 1.0.7
1076
1077- Renamed /allowed-tools -> /permissions
1078- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1079- Deprecated claude config commands in favor of editing settings.json
1080- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1081- Improved error handling for /install-github-app
1082- Bugfixes, UI polish, and tool reliability improvements
1083
1084## 1.0.6
1085
1086- Improved edit reliability for tab-indented files
1087- Respect CLAUDE_CONFIG_DIR everywhere
1088- Reduced unnecessary tool permission prompts
1089- Added support for symlinks in @file typeahead
1090- Bugfixes, UI polish, and tool reliability improvements
1091
1092## 1.0.4
1093
1094- Fixed a bug where MCP tool errors weren't being parsed correctly
1095
1096## 1.0.1
1097
1098- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1099- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1100- Updated documentation links and OAuth process descriptions
1101
1102## 1.0.0
1103
1104- Claude Code is now generally available
1105- Introducing Sonnet 4 and Opus 4 models
1106
1107## 0.2.125
1108
1109- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1110- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1111
1112## 0.2.117
1113
1114- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1115- Introduced settings.cleanupPeriodDays
1116- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1117- Introduced --debug mode
1118
1119## 0.2.108
1120
1121- You can now send messages to Claude while it works to steer Claude in real-time
1122- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1123- Fixed a bug where thinking was not working in -p mode
1124- Fixed a regression in /cost reporting
1125- Deprecated MCP wizard interface in favor of other MCP commands
1126- Lots of other bugfixes and improvements
1127
1128## 0.2.107
1129
1130- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1131
1132## 0.2.106
1133
1134- MCP SSE server configs can now specify custom headers
1135- Fixed a bug where MCP permission prompt didn't always show correctly
1136
1137## 0.2.105
1138
1139- Claude can now search the web
1140- Moved system & account status to /status
1141- Added word movement keybindings for Vim
1142- Improved latency for startup, todo tool, and file edits
1143
1144## 0.2.102
1145
1146- Improved thinking triggering reliability
1147- Improved @mention reliability for images and folders
1148- You can now paste multiple large chunks into one prompt
1149
1150## 0.2.100
1151
1152- Fixed a crash caused by a stack overflow error
1153- Made db storage optional; missing db support disables --continue and --resume
1154
1155## 0.2.98
1156
1157- Fixed an issue where auto-compact was running twice
1158
1159## 0.2.96
1160
1161- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1162
1163## 0.2.93
1164
1165- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1166- Claude now has access to a Todo list that helps it stay on track and be more organized
1167
1168## 0.2.82
1169
1170- Added support for --disallowedTools
1171- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1172
1173## 0.2.75
1174
1175- Hit Enter to queue up additional messages while Claude is working
1176- Drag in or copy/paste image files directly into the prompt
1177- @-mention files to directly add them to context
1178- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1179- Improved performance for filename auto-complete
1180
1181## 0.2.74
1182
1183- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1184- Task tool can now perform writes and run bash commands
1185
1186## 0.2.72
1187
1188- Updated spinner to indicate tokens loaded and tool usage
1189
1190## 0.2.70
1191
1192- Network commands like curl are now available for Claude to use
1193- Claude can now run multiple web queries in parallel
1194- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1195
1196## 0.2.69
1197
1198- Fixed UI glitches with improved Select component behavior
1199- Enhanced terminal output display with better text truncation logic
1200
1201## 0.2.67
1202
1203- Shared project permission rules can be saved in .claude/settings.json
1204
1205## 0.2.66
1206
1207- Print mode (-p) now supports streaming output via --output-format=stream-json
1208- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1209
1210## 0.2.63
1211
1212- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1213
1214## 0.2.61
1215
1216- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1217- Enhanced image detection for more reliable clipboard paste functionality
1218- Fixed an issue where ESC key could crash the conversation history selector
1219
1220## 0.2.59
1221
1222- Copy+paste images directly into your prompt
1223- Improved progress indicators for bash and fetch tools
1224- Bugfixes for non-interactive mode (-p)
1225
1226## 0.2.54
1227
1228- Quickly add to Memory by starting your message with '#'
1229- Press ctrl+r to see full output for long tool results
1230- Added support for MCP SSE transport
1231
1232## 0.2.53
1233
1234- New web fetch tool lets Claude view URLs that you paste in
1235- Fixed a bug with JPEG detection
1236
1237## 0.2.50
1238
1239- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1240
1241## 0.2.49
1242
1243- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1244
1245## 0.2.47
1246
1247- Press Tab to auto-complete file and folder names
1248- Press Shift + Tab to toggle auto-accept for file edits
1249- Automatic conversation compaction for infinite conversation length (toggle with /config)
1250
1251## 0.2.44
1252
1253- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1254
1255## 0.2.41
1256
1257- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1258- MCP server startup no longer blocks the app from starting up
1259
1260## 0.2.37
1261
1262- New /release-notes command lets you view release notes at any time
1263- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1264
1265## 0.2.36
1266
1267- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1268- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1269
1270## 0.2.34
1271
1272- Vim bindings for text input - enable with /vim or /config
1273
1274## 0.2.32
1275
1276- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1277- Fix for some PersistentShell issues
1278
1279## 0.2.31
1280
1281- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1282- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1283
1284## 0.2.30
1285
1286- Added ANSI color theme for better terminal compatibility
1287- Fixed issue where slash command arguments weren't being sent properly
1288- (Mac-only) API keys are now stored in macOS Keychain
1289
1290## 0.2.26
1291
1292- New /approved-tools command for managing tool permissions
1293- Word-level diff display for improved code readability
1294- Fuzzy matching for slash commands
1295
1296## 0.2.21
1297
1298- Fuzzy matching for /commands