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 d6ab0ea · on Feb 26, 2026 History
1,915 lines · 110.8 KBPreviewCodeRaw
1# Changelog
2
3## 2.1.59
4
5- Claude automatically saves useful context to auto-memory. Manage with /memory
6- Added `/copy` command to show an interactive picker when code blocks are present, allowing selection of individual code blocks or the full response.
7- Improved "always allow" prefix suggestions for compound bash commands (e.g. `cd /tmp && git fetch && git push`) to compute smarter per-subcommand prefixes instead of treating the whole command as one
8- Improved ordering of short task lists
9- Improved memory usage in multi-agent sessions by releasing completed subagent task state
10- Fixed MCP OAuth token refresh race condition when running multiple Claude Code instances simultaneously
11- Fixed shell commands not showing a clear error message when the working directory has been deleted
12
13## 2.1.58
14
15- Expand Remote Control to more users
16
17## 2.1.56
18
19- VS Code: Fixed another cause of "command 'claude-vscode.editor.openLast' not found" crashes
20
21## 2.1.55
22
23- Fixed BashTool failing on Windows with EINVAL error
24
25## 2.1.53
26
27- Fixed a UI flicker where user input would briefly disappear after submission before the message rendered
28- Fixed bulk agent kill (ctrl+f) to send a single aggregate notification instead of one per agent, and to properly clear the command queue
29- Fixed graceful shutdown sometimes leaving stale sessions when using Remote Control by parallelizing teardown network calls
30- Fixed `--worktree` sometimes being ignored on first launch
31- Fixed a panic ("switch on corrupted value") on Windows
32- Fixed a crash that could occur when spawning many processes on Windows
33- Fixed a crash in the WebAssembly interpreter on Linux x64 & Windows x64
34- Fixed a crash that sometimes occurred after 2 minutes on Windows ARM64
35
36## 2.1.52
37
38- VS Code: Fixed extension crash on Windows ("command 'claude-vscode.editor.openLast' not found")
39
40## 2.1.51
41
42- Added `claude remote-control` subcommand for external builds, enabling local environment serving for all users.
43- Updated plugin marketplace default git timeout from 30s to 120s and added `CLAUDE_CODE_PLUGIN_GIT_TIMEOUT_MS` to configure.
44- Added support for custom npm registries and specific version pinning when installing plugins from npm sources
45- BashTool now skips login shell (`-l` flag) by default when a shell snapshot is available, improving command execution performance. Previously this required setting `CLAUDE_BASH_NO_LOGIN=true`.
46- Fixed a security issue where `statusLine` and `fileSuggestion` hook commands could execute without workspace trust acceptance in interactive mode.
47- Tool results larger than 50K characters are now persisted to disk (previously 100K). This reduces context window usage and improves conversation longevity.
48- Fixed a bug where duplicate `control_response` messages (e.g. from WebSocket reconnects) could cause API 400 errors by pushing duplicate assistant messages into the conversation.
49- Added `CLAUDE_CODE_ACCOUNT_UUID`, `CLAUDE_CODE_USER_EMAIL`, and `CLAUDE_CODE_ORGANIZATION_UUID` environment variables for SDK callers to provide account info synchronously, eliminating a race condition where early telemetry events lacked account metadata.
50- Fixed slash command autocomplete crashing when a plugin's SKILL.md description is a YAML array or other non-string type
51- The `/model` picker now shows human-readable labels (e.g., "Sonnet 4.5") instead of raw model IDs for pinned model versions, with an upgrade hint when a newer version is available.
52- Managed settings can now be set via macOS plist or Windows Registry. Learn more at https://code.claude.com/docs/en/settings#settings-files
53
54## 2.1.50
55
56- Added support for `startupTimeout` configuration for LSP servers
57- Added `WorktreeCreate` and `WorktreeRemove` hook events, enabling custom VCS setup and teardown when agent worktree isolation creates or removes worktrees.
58- Fixed a bug where resumed sessions could be invisible when the working directory involved symlinks, because the session storage path was resolved at different times during startup. Also fixed session data loss on SSH disconnect by flushing session data before hooks and analytics in the graceful shutdown sequence.
59- Linux: Fixed native modules not loading on systems with glibc older than 2.30 (e.g., RHEL 8)
60- Fixed memory leak in agent teams where completed teammate tasks were never garbage collected from session state
61- Fixed `CLAUDE_CODE_SIMPLE` to fully strip down skills, session memory, custom agents, and CLAUDE.md token counting
62- Fixed `/mcp reconnect` freezing the CLI when given a server name that doesn't exist
63- Fixed memory leak where completed task state objects were never removed from AppState
64- Added support for `isolation: worktree` in agent definitions, allowing agents to declaratively run in isolated git worktrees.
65- `CLAUDE_CODE_SIMPLE` mode now also disables MCP tools, attachments, hooks, and CLAUDE.md file loading for a fully minimal experience.
66- Fixed bug where MCP tools were not discovered when tool search is enabled and a prompt is passed in as a launch argument
67- Improved memory usage during long sessions by clearing internal caches after compaction
68- Added `claude agents` CLI command to list all configured agents
69- Improved memory usage during long sessions by clearing large tool results after they have been processed
70- Fixed a memory leak where LSP diagnostic data was never cleaned up after delivery, causing unbounded memory growth in long sessions
71- Fixed a memory leak where completed task output was not freed from memory, reducing memory usage in long sessions with many tasks
72- Improved startup performance for headless mode (`-p` flag) by deferring Yoga WASM and UI component imports
73- Fixed prompt suggestion cache regression that reduced cache hit rates
74- Fixed unbounded memory growth in long sessions by capping file history snapshots
75- Added `CLAUDE_CODE_DISABLE_1M_CONTEXT` environment variable to disable 1M context window support
76- Opus 4.6 (fast mode) now includes the full 1M context window
77- VSCode: Added `/extra-usage` command support in VS Code sessions
78- Fixed memory leak where TaskOutput retained recent lines after cleanup
79- Fixed memory leak in CircularBuffer where cleared items were retained in the backing array
80- Fixed memory leak in shell command execution where ChildProcess and AbortController references were retained after cleanup
81
82## 2.1.49
83
84- Improved MCP OAuth authentication with step-up auth support and discovery caching, reducing redundant network requests during server connections
85- Added `--worktree` (`-w`) flag to start Claude in an isolated git worktree
86- Subagents support `isolation: "worktree"` for working in a temporary git worktree
87- Added Ctrl+F keybinding to kill background agents (two-press confirmation)
88- Agent definitions support `background: true` to always run as a background task
89- Plugins can ship `settings.json` for default configuration
90- Fixed file-not-found errors to suggest corrected paths when the model drops the repo folder
91- Fixed Ctrl+C and ESC being silently ignored when background agents are running and the main thread is idle. Pressing twice within 3 seconds now kills all background agents.
92- Fixed prompt suggestion cache regression that reduced cache hit rates.
93- Fixed `plugin enable` and `plugin disable` to auto-detect the correct scope when `--scope` is not specified, instead of always defaulting to user scope
94- Simple mode (`CLAUDE_CODE_SIMPLE`) now includes the file edit tool in addition to the Bash tool, allowing direct file editing in simple mode.
95- Permission suggestions are now populated when safety checks trigger an ask response, enabling SDK consumers to display permission options
96- Sonnet 4.5 with 1M context is being removed from the Max plan in favor of our frontier Sonnet 4.6 model, which now has 1M context. Please switch in /model.
97- Fixed verbose mode not updating thinking block display when toggled via `/config` — memo comparators now correctly detect verbose changes
98- Fixed unbounded WASM memory growth during long sessions by periodically resetting the tree-sitter parser
99- Fixed potential rendering issues caused by stale yoga layout references
100- Improved performance in non-interactive mode (`-p`) by skipping unnecessary API calls during startup
101- Improved performance by caching authentication failures for HTTP and SSE MCP servers, avoiding repeated connection attempts to servers requiring auth
102- Fixed unbounded memory growth during long-running sessions caused by Yoga WASM linear memory never shrinking
103- SDK model info now includes `supportsEffort`, `supportedEffortLevels`, and `supportsAdaptiveThinking` fields so consumers can discover model capabilities.
104- Added `ConfigChange` hook event that fires when configuration files change during a session, enabling enterprise security auditing and optional blocking of settings changes.
105- Improved startup performance by caching MCP auth failures to avoid redundant connection attempts
106- Improved startup performance by reducing HTTP calls for analytics token counting
107- Improved startup performance by batching MCP tool token counting into a single API call
108- Fixed `disableAllHooks` setting to respect managed settings hierarchy — non-managed settings can no longer disable managed hooks set by policy (#26637)
109- Fixed `--resume` session picker showing raw XML tags for sessions that start with commands like `/clear`. Now correctly falls through to the session ID fallback.
110- Improved permission prompts for path safety and working directory blocks to show the reason for the restriction instead of a bare prompt with no context
111
112## 2.1.47
113
114- Fixed FileWriteTool line counting to preserve intentional trailing blank lines instead of stripping them with `trimEnd()`.
115- Fixed Windows terminal rendering bugs caused by `os.EOL` (`\r\n`) in display code — line counts now show correct values instead of always showing 1 on Windows.
116- Improved VS Code plan preview: auto-updates as Claude iterates, enables commenting only when the plan is ready for review, and keeps the preview open when rejecting so Claude can revise.
117- Fixed a bug where bold and colored text in markdown output could shift to the wrong characters on Windows due to `\r\n` line endings.
118- Fixed compaction failing when conversation contains many PDF documents by stripping document blocks alongside images before sending to the compaction API (anthropics/claude-code#26188)
119- Improved memory usage in long-running sessions by releasing API stream buffers, agent context, and skill state after use
120- Improved startup performance by deferring SessionStart hook execution, reducing time-to-interactive by ~500ms.
121- Fixed an issue where bash tool output was silently discarded on Windows when using MSYS2 or Cygwin shells.
122- Improved performance of `@` file mentions - file suggestions now appear faster by pre-warming the index on startup and using session-based caching with background refresh.
123- Improved memory usage by trimming agent task message history after tasks complete
124- Improved memory usage during long agent sessions by eliminating O(n²) message accumulation in progress updates
125- Fixed the bash permission classifier to validate that returned match descriptions correspond to actual input rules, preventing hallucinated descriptions from incorrectly granting permissions
126- Fixed user-defined agents only loading one file on NFS/FUSE filesystems that report zero inodes (anthropics/claude-code#26044)
127- Fixed plugin agent skills silently failing to load when referenced by bare name instead of fully-qualified plugin name (anthropics/claude-code#25834)
128- Search patterns in collapsed tool results are now displayed in quotes for clarity
129- Windows: Fixed CWD tracking temp files never being cleaned up, causing them to accumulate indefinitely (anthropics/claude-code#17600)
130- Use `ctrl+f` to kill all background agents instead of double-pressing ESC. Background agents now continue running when you press ESC to cancel the main thread, giving you more control over agent lifecycle.
131- Fixed API 400 errors ("thinking blocks cannot be modified") that occurred in sessions with concurrent agents, caused by interleaved streaming content blocks preventing proper message merging.
132- Simplified teammate navigation to use only Shift+Down (with wrapping) instead of both Shift+Up and Shift+Down.
133- Fixed an issue where a single file write/edit error would abort all other parallel file write/edit operations. Independent file mutations now complete even when a sibling fails.
134- Added `last_assistant_message` field to Stop and SubagentStop hook inputs, providing the final assistant response text so hooks can access it without parsing transcript files.
135- Fixed custom session titles set via `/rename` being lost after resuming a conversation (anthropics/claude-code#23610)
136- Fixed collapsed read/search hint text overflowing on narrow terminals by truncating from the start.
137- Fixed an issue where bash commands with backslash-newline continuation lines (e.g., long commands split across multiple lines with `\`) would produce spurious empty arguments, potentially breaking command execution.
138- Fixed built-in slash commands (`/help`, `/model`, `/compact`, etc.) being hidden from the autocomplete dropdown when many user skills are installed (anthropics/claude-code#22020)
139- Fixed MCP servers not appearing in the MCP Management Dialog after deferred loading
140- Fixed session name persisting in status bar after `/clear` command (anthropics/claude-code#26082)
141- Fixed crash when a skill's `name` or `description` in SKILL.md frontmatter is a bare number (e.g., `name: 3000`) — the value is now properly coerced to a string (anthropics/claude-code#25837)
142- Fixed /resume silently dropping sessions when the first message exceeds 16KB or uses array-format content (anthropics/claude-code#25721)
143- Added `chat:newline` keybinding action for configurable multi-line input (anthropics/claude-code#26075)
144- Added `added_dirs` to the statusline JSON `workspace` section, exposing directories added via `/add-dir` to external scripts (anthropics/claude-code#26096)
145- Fixed `claude doctor` misclassifying mise and asdf-managed installations as native installs (anthropics/claude-code#26033)
146- Fixed zsh heredoc failing with "read-only file system" error in sandboxed commands (anthropics/claude-code#25990)
147- Fixed agent progress indicator showing inflated tool use count (anthropics/claude-code#26023)
148- Fixed image pasting not working on WSL2 systems where Windows copies images as BMP format (anthropics/claude-code#25935)
149- Fixed background agent results returning raw transcript data instead of the agent's final answer (anthropics/claude-code#26012)
150- Fixed Warp terminal incorrectly prompting for Shift+Enter setup when it supports it natively (anthropics/claude-code#25957)
151- Fixed CJK wide characters causing misaligned timestamps and layout elements in the TUI (anthropics/claude-code#26084)
152- Fixed custom agent `model` field in `.claude/agents/*.md` being ignored when spawning team teammates (anthropics/claude-code#26064)
153- Fixed plan mode being lost after context compaction, causing the model to switch from planning to implementation mode (anthropics/claude-code#26061)
154- Fixed `alwaysThinkingEnabled: true` in settings.json not enabling thinking mode on Bedrock and Vertex providers (anthropics/claude-code#26074)
155- Fixed `tool_decision` OTel telemetry event not being emitted in headless/SDK mode (anthropics/claude-code#26059)
156- Fixed session name being lost after context compaction — renamed sessions now preserve their custom title through compaction (anthropics/claude-code#26121)
157- Increased initial session count in resume picker from 10 to 50 for faster session discovery (anthropics/claude-code#26123)
158- Windows: fixed worktree session matching when drive letter casing differs (anthropics/claude-code#26123)
159- Fixed `/resume <session-id>` failing to find sessions whose first message exceeds 16KB (anthropics/claude-code#25920)
160- Fixed "Always allow" on multiline bash commands creating invalid permission patterns that corrupt settings (anthropics/claude-code#25909)
161- Fixed React crash (error #31) when a skill's `argument-hint` in SKILL.md frontmatter uses YAML sequence syntax (e.g., `[topic: foo | bar]`) — the value is now properly coerced to a string (anthropics/claude-code#25826)
162- Fixed crash when using `/fork` on sessions that used web search — null entries in search results from transcript deserialization are now handled gracefully (anthropics/claude-code#25811)
163- Fixed read-only git commands triggering FSEvents file watcher loops on macOS by adding --no-optional-locks flag (anthropics/claude-code#25750)
164- Fixed custom agents and skills not being discovered when running from a git worktree — project-level `.claude/agents/` and `.claude/skills/` from the main repository are now included (anthropics/claude-code#25816)
165- Fixed non-interactive subcommands like `claude doctor` and `claude plugin validate` being blocked inside nested Claude sessions (anthropics/claude-code#25803)
166- Windows: Fixed the same CLAUDE.md file being loaded twice when drive letter casing differs between paths (anthropics/claude-code#25756)
167- Fixed inline code spans in markdown being incorrectly parsed as bash commands (anthropics/claude-code#25792)
168- Fixed teammate spinners not respecting custom spinnerVerbs from settings (anthropics/claude-code#25748)
169- Fixed shell commands permanently failing after a command deletes its own working directory (anthropics/claude-code#26136)
170- Fixed hooks (PreToolUse, PostToolUse) silently failing to execute on Windows by using Git Bash instead of cmd.exe (anthropics/claude-code#25981)
171- Fixed LSP `findReferences` and other location-based operations returning results from gitignored files (e.g., `node_modules/`, `venv/`) (anthropics/claude-code#26051)
172- Moved config backup files from home directory root to `~/.claude/backups/` to reduce home directory clutter (anthropics/claude-code#26130)
173- Fixed sessions with large first prompts (>16KB) disappearing from the /resume list (anthropics/claude-code#26140)
174- Fixed shell functions with double-underscore prefixes (e.g., `__git_ps1`) not being preserved across shell sessions (anthropics/claude-code#25824)
175- Fixed spinner showing "0 tokens" counter before any tokens have been received (anthropics/claude-code#26105)
176- VSCode: Fixed conversation messages appearing dimmed while the AskUserQuestion dialog is open (anthropics/claude-code#26078)
177- Fixed background tasks failing in git worktrees due to remote URL resolution reading from worktree-specific gitdir instead of the main repository config (anthropics/claude-code#26065)
178- Fixed Right Alt key leaving visible `[25~` escape sequence residue in the input field on Windows/Git Bash terminals (anthropics/claude-code#25943)
179- The `/rename` command now updates the terminal tab title by default (anthropics/claude-code#25789)
180- Fixed Edit tool silently corrupting Unicode curly quotes (\u201c\u201d \u2018\u2019) by replacing them with straight quotes when making edits (anthropics/claude-code#26141)
181- Fixed OSC 8 hyperlinks only being clickable on the first line when link text wraps across multiple terminal lines.
182
183## 2.1.46
184
185- Fixed orphaned CC processes after terminal disconnect on macOS
186- Added support for using claude.ai MCP connectors in Claude Code
187
188## 2.1.45
189
190- Added support for Claude Sonnet 4.6
191- Added support for reading `enabledPlugins` and `extraKnownMarketplaces` from `--add-dir` directories
192- Added `spinnerTipsOverride` setting to customize spinner tips — configure `tips` with an array of custom tip strings, and optionally set `excludeDefault: true` to show only your custom tips instead of the built-in ones
193- Added `SDKRateLimitInfo` and `SDKRateLimitEvent` types to the SDK, enabling consumers to receive rate limit status updates including utilization, reset times, and overage information
194- Fixed Agent Teams teammates failing on Bedrock, Vertex, and Foundry by propagating API provider environment variables to tmux-spawned processes (anthropics/claude-code#23561)
195- Fixed sandbox "operation not permitted" errors when writing temporary files on macOS by using the correct per-user temp directory (anthropics/claude-code#21654)
196- Fixed Task tool (backgrounded agents) crashing with a `ReferenceError` on completion (anthropics/claude-code#22087)
197- Fixed autocomplete suggestions not being accepted on Enter when images are pasted in the input
198- Fixed skills invoked by subagents incorrectly appearing in main session context after compaction
199- Fixed excessive `.claude.json.backup` files accumulating on every startup
200- Fixed plugin-provided commands, agents, and hooks not being available immediately after installation without requiring a restart
201- Improved startup performance by removing eager loading of session history for stats caching
202- Improved memory usage for shell commands that produce large output — RSS no longer grows unboundedly with command output size
203- Improved collapsed read/search groups to show the current file or search pattern being processed beneath the summary line while active
204- [VSCode] Improved permission destination choice (project/user/session) to persist across sessions
205
206## 2.1.44
207
208- Fixed ENAMETOOLONG errors for deeply-nested directory paths
209- Fixed auth refresh errors
210
211## 2.1.43
212
213- Fixed AWS auth refresh hanging indefinitely by adding a 3-minute timeout
214- Fixed spurious warnings for non-agent markdown files in `.claude/agents/` directory
215- Fixed structured-outputs beta header being sent unconditionally on Vertex/Bedrock
216
217## 2.1.42
218
219- Improved startup performance by deferring Zod schema construction
220- Improved prompt cache hit rates by moving date out of system prompt
221- Added one-time Opus 4.6 effort callout for eligible users
222- Fixed /resume showing interrupt messages as session titles
223- Fixed image dimension limit errors to suggest /compact
224
225## 2.1.41
226
227- Added guard against launching Claude Code inside another Claude Code session
228- Fixed Agent Teams using wrong model identifier for Bedrock, Vertex, and Foundry customers
229- Fixed a crash when MCP tools return image content during streaming
230- Fixed /resume session previews showing raw XML tags instead of readable command names
231- Improved model error messages for Bedrock/Vertex/Foundry users with fallback suggestions
232- Fixed plugin browse showing misleading "Space to Toggle" hint for already-installed plugins
233- Fixed hook blocking errors (exit code 2) not showing stderr to the user
234- Added `speed` attribute to OTel events and trace spans for fast mode visibility
235- Added `claude auth login`, `claude auth status`, and `claude auth logout` CLI subcommands
236- Added Windows ARM64 (win32-arm64) native binary support
237- Improved `/rename` to auto-generate session name from conversation context when called without arguments
238- Improved narrow terminal layout for prompt footer
239- Fixed file resolution failing for @-mentions with anchor fragments (e.g., `@README.md#installation`)
240- Fixed FileReadTool blocking the process on FIFOs, `/dev/stdin`, and large files
241- Fixed background task notifications not being delivered in streaming Agent SDK mode
242- Fixed cursor jumping to end on each keystroke in classifier rule input
243- Fixed markdown link display text being dropped for raw URL
244- Fixed auto-compact failure error notifications being shown to users
245- Fixed permission wait time being included in subagent elapsed time display
246- Fixed proactive ticks firing while in plan mode
247- Fixed clear stale permission rules when settings change on disk
248- Fixed hook blocking errors showing stderr content in UI
249
250## 2.1.39
251
252- Improved terminal rendering performance
253- Fixed fatal errors being swallowed instead of displayed
254- Fixed process hanging after session close
255- Fixed character loss at terminal screen boundary
256- Fixed blank lines in verbose transcript view
257
258## 2.1.38
259
260- Fixed VS Code terminal scroll-to-top regression introduced in 2.1.37
261- Fixed Tab key queueing slash commands instead of autocompleting
262- Fixed bash permission matching for commands using environment variable wrappers
263- Fixed text between tool uses disappearing when not using streaming
264- Fixed duplicate sessions when resuming in VS Code extension
265- Improved heredoc delimiter parsing to prevent command smuggling
266- Blocked writes to `.claude/skills` directory in sandbox mode
267
268## 2.1.37
269
270- Fixed an issue where /fast was not immediately available after enabling /extra-usage
271
272## 2.1.36
273
274- Fast mode is now available for Opus 4.6. Learn more at https://code.claude.com/docs/en/fast-mode
275
276## 2.1.34
277
278- Fixed a crash when agent teams setting changed between renders
279- Fixed a bug where commands excluded from sandboxing (via `sandbox.excludedCommands` or `dangerouslyDisableSandbox`) could bypass the Bash ask permission rule when `autoAllowBashIfSandboxed` was enabled
280
281## 2.1.33
282
283- Fixed agent teammate sessions in tmux to send and receive messages
284- Fixed warnings about agent teams not being available on your current plan
285- Added `TeammateIdle` and `TaskCompleted` hook events for multi-agent workflows
286- Added support for restricting which sub-agents can be spawned via `Task(agent_type)` syntax in agent "tools" frontmatter
287- Added `memory` frontmatter field support for agents, enabling persistent memory with `user`, `project`, or `local` scope
288- Added plugin name to skill descriptions and `/skills` menu for better discoverability
289- Fixed an issue where submitting a new message while the model was in extended thinking would interrupt the thinking phase
290- Fixed an API error that could occur when aborting mid-stream, where whitespace text combined with a thinking block would bypass normalization and produce an invalid request
291- Fixed API proxy compatibility issue where 404 errors on streaming endpoints no longer triggered non-streaming fallback
292- Fixed an issue where proxy settings configured via `settings.json` environment variables were not applied to WebFetch and other HTTP requests on the Node.js build
293- Fixed `/resume` session picker showing raw XML markup instead of clean titles for sessions started with slash commands
294- Improved error messages for API connection failures — now shows specific cause (e.g., ECONNREFUSED, SSL errors) instead of generic "Connection error"
295- Errors from invalid managed settings are now surfaced
296- VSCode: Added support for remote sessions, allowing OAuth users to browse and resume sessions from claude.ai
297- VSCode: Added git branch and message count to the session picker, with support for searching by branch name
298- VSCode: Fixed scroll-to-bottom under-scrolling on initial session load and session switch
299
300## 2.1.32
301
302- Claude Opus 4.6 is now available!
303- Added research preview agent teams feature for multi-agent collaboration (token-intensive feature, requires setting CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1)
304- Claude now automatically records and recalls memories as it works
305- Added "Summarize from here" to the message selector, allowing partial conversation summarization.
306- Skills defined in `.claude/skills/` within additional directories (`--add-dir`) are now loaded automatically.
307- Fixed `@` file completion showing incorrect relative paths when running from a subdirectory
308- Updated --resume to re-use --agent value specified in previous conversation by default.
309- Fixed: Bash tool no longer throws "Bad substitution" errors when heredocs contain JavaScript template literals like `${index + 1}`, which previously interrupted tool execution
310- Skill character budget now scales with context window (2% of context), so users with larger context windows can see more skill descriptions without truncation
311- Fixed Thai/Lao spacing vowels (สระ า, ำ) not rendering correctly in the input field
312- VSCode: Fixed slash commands incorrectly being executed when pressing Enter with preceding text in the input field
313- VSCode: Added spinner when loading past conversations list
314
315## 2.1.31
316
317- Added session resume hint on exit, showing how to continue your conversation later
318- Added support for full-width (zenkaku) space input from Japanese IME in checkbox selection
319- Fixed PDF too large errors permanently locking up sessions, requiring users to start a new conversation
320- Fixed bash commands incorrectly reporting failure with "Read-only file system" errors when sandbox mode was enabled
321- Fixed a crash that made sessions unusable after entering plan mode when project config in `~/.claude.json` was missing default fields
322- Fixed `temperatureOverride` being silently ignored in the streaming API path, causing all streaming requests to use the default temperature (1) regardless of the configured override
323- Fixed LSP shutdown/exit compatibility with strict language servers that reject null params
324- Improved system prompts to more clearly guide the model toward using dedicated tools (Read, Edit, Glob, Grep) instead of bash equivalents (`cat`, `sed`, `grep`, `find`), reducing unnecessary bash command usage
325- Improved PDF and request size error messages to show actual limits (100 pages, 20MB)
326- Reduced layout jitter in the terminal when the spinner appears and disappears during streaming
327- Removed misleading Anthropic API pricing from model selector for third-party provider (Bedrock, Vertex, Foundry) users
328
329## 2.1.30
330
331- Added `pages` parameter to the Read tool for PDFs, allowing specific page ranges to be read (e.g., `pages: "1-5"`). Large PDFs (>10 pages) now return a lightweight reference when `@` mentioned instead of being inlined into context.
332- Added pre-configured OAuth client credentials for MCP servers that don't support Dynamic Client Registration (e.g., Slack). Use `--client-id` and `--client-secret` with `claude mcp add`.
333- Added `/debug` for Claude to help troubleshoot the current session
334- Added support for additional `git log` and `git show` flags in read-only mode (e.g., `--topo-order`, `--cherry-pick`, `--format`, `--raw`)
335- Added token count, tool uses, and duration metrics to Task tool results
336- Added reduced motion mode to the config
337- Fixed phantom "(no content)" text blocks appearing in API conversation history, reducing token waste and potential model confusion
338- Fixed prompt cache not correctly invalidating when tool descriptions or input schemas changed, only when tool names changed
339- Fixed 400 errors that could occur after running `/login` when the conversation contained thinking blocks
340- Fixed a hang when resuming sessions with corrupted transcript files containing `parentUuid` cycles
341- Fixed rate limit message showing incorrect "/upgrade" suggestion for Max 20x users when extra-usage is unavailable
342- Fixed permission dialogs stealing focus while actively typing
343- Fixed subagents not being able to access SDK-provided MCP tools because they were not synced to the shared application state
344- Fixed a regression where Windows users with a `.bashrc` file could not run bash commands
345- Improved memory usage for `--resume` (68% reduction for users with many sessions) by replacing the session index with lightweight stat-based loading and progressive enrichment
346- Improved `TaskStop` tool to display the stopped command/task description in the result line instead of a generic "Task stopped" message
347- Changed `/model` to execute immediately instead of being queued
348- [VSCode] Added multiline input support to the "Other" text input in question dialogs (use Shift+Enter for new lines)
349- [VSCode] Fixed duplicate sessions appearing in the session list when starting a new conversation
350
351## 2.1.29
352
353- Fixed startup performance issues when resuming sessions that have `saved_hook_context`
354
355## 2.1.27
356
357- Added tool call failures and denials to debug logs
358- Fixed context management validation error for gateway users, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
359- Added `--from-pr` flag to resume sessions linked to a specific GitHub PR number or URL
360- Sessions are now automatically linked to PRs when created via `gh pr create`
361- Fixed /context command not displaying colored output
362- Fixed status bar duplicating background task indicator when PR status was shown
363- Windows: Fixed bash command execution failing for users with `.bashrc` files
364- Windows: Fixed console windows flashing when spawning child processes
365- VSCode: Fixed OAuth token expiration causing 401 errors after extended sessions
366
367## 2.1.25
368
369- Fixed beta header validation error for gateway users on Bedrock and Vertex, ensuring `CLAUDE_CODE_DISABLE_EXPERIMENTAL_BETAS=1` avoids the error
370
371## 2.1.23
372
373- Added customizable spinner verbs setting (`spinnerVerbs`)
374- Fixed mTLS and proxy connectivity for users behind corporate proxies or using client certificates
375- Fixed per-user temp directory isolation to prevent permission conflicts on shared systems
376- Fixed a race condition that could cause 400 errors when prompt caching scope was enabled
377- Fixed pending async hooks not being cancelled when headless streaming sessions ended
378- Fixed tab completion not updating the input field when accepting a suggestion
379- Fixed ripgrep search timeouts silently returning empty results instead of reporting errors
380- Improved terminal rendering performance with optimized screen data layout
381- Changed Bash commands to show timeout duration alongside elapsed time
382- Changed merged pull requests to show a purple status indicator in the prompt footer
383- [IDE] Fixed model options displaying incorrect region strings for Bedrock users in headless mode
384
385## 2.1.22
386
387- Fixed structured outputs for non-interactive (-p) mode
388
389## 2.1.21
390
391- Added support for full-width (zenkaku) number input from Japanese IME in option selection prompts
392- Fixed shell completion cache files being truncated on exit
393- Fixed API errors when resuming sessions that were interrupted during tool execution
394- Fixed auto-compact triggering too early on models with large output token limits
395- Fixed task IDs potentially being reused after deletion
396- Fixed file search not working in VS Code extension on Windows
397- Improved read/search progress indicators to show "Reading…" while in progress and "Read" when complete
398- Improved Claude to prefer file operation tools (Read, Edit, Write) over bash equivalents (cat, sed, awk)
399- [VSCode] Added automatic Python virtual environment activation, ensuring `python` and `pip` commands use the correct interpreter (configurable via `claudeCode.usePythonEnvironment` setting)
400- [VSCode] Fixed message action buttons having incorrect background colors
401
402## 2.1.20
403
404- Added arrow key history navigation in vim normal mode when cursor cannot move further
405- Added external editor shortcut (Ctrl+G) to the help menu for better discoverability
406- Added PR review status indicator to the prompt footer, showing the current branch's PR state (approved, changes requested, pending, or draft) as a colored dot with a clickable link
407- Added support for loading `CLAUDE.md` files from additional directories specified via `--add-dir` flag (requires setting `CLAUDE_CODE_ADDITIONAL_DIRECTORIES_CLAUDE_MD=1`)
408- Added ability to delete tasks via the `TaskUpdate` tool
409- Fixed session compaction issues that could cause resume to load full history instead of the compact summary
410- Fixed agents sometimes ignoring user messages sent while actively working on a task
411- Fixed wide character (emoji, CJK) rendering artifacts where trailing columns were not cleared when replaced by narrower characters
412- Fixed JSON parsing errors when MCP tool responses contain special Unicode characters
413- Fixed up/down arrow keys in multi-line and wrapped text input to prioritize cursor movement over history navigation
414- Fixed draft prompt being lost when pressing UP arrow to navigate command history
415- Fixed ghost text flickering when typing slash commands mid-input
416- Fixed marketplace source removal not properly deleting settings
417- Fixed duplicate output in some commands like `/context`
418- Fixed task list sometimes showing outside the main conversation view
419- Fixed syntax highlighting for diffs occurring within multiline constructs like Python docstrings
420- Fixed crashes when cancelling tool use
421- Improved `/sandbox` command UI to show dependency status with installation instructions when dependencies are missing
422- Improved thinking status text with a subtle shimmer animation
423- Improved task list to dynamically adjust visible items based on terminal height
424- Improved fork conversation hint to show how to resume the original session
425- Changed collapsed read/search groups to show present tense ("Reading", "Searching for") while in progress, and past tense ("Read", "Searched for") when complete
426- Changed `ToolSearch` results to appear as a brief notification instead of inline in the conversation
427- Changed the `/commit-push-pr` skill to automatically post PR URLs to Slack channels when configured via MCP tools
428- Changed the `/copy` command to be available to all users
429- Changed background agents to prompt for tool permissions before launching
430- Changed permission rules like `Bash(*)` to be accepted and treated as equivalent to `Bash`
431- Changed config backups to be timestamped and rotated (keeping 5 most recent) to prevent data loss
432
433## 2.1.19
434
435- Added env var `CLAUDE_CODE_ENABLE_TASKS`, set to `false` to keep the old system temporarily
436- Added shorthand `$0`, `$1`, etc. for accessing individual arguments in custom commands
437- Fixed crashes on processors without AVX instruction support
438- Fixed dangling Claude Code processes when terminal is closed by catching EIO errors from `process.exit()` and using SIGKILL as fallback
439- Fixed `/rename` and `/tag` not updating the correct session when resuming from a different directory (e.g., git worktrees)
440- Fixed resuming sessions by custom title not working when run from a different directory
441- Fixed pasted text content being lost when using prompt stash (Ctrl+S) and restore
442- Fixed agent list displaying "Sonnet (default)" instead of "Inherit (default)" for agents without an explicit model setting
443- Fixed backgrounded hook commands not returning early, potentially causing the session to wait on a process that was intentionally backgrounded
444- Fixed file write preview omitting empty lines
445- Changed skills without additional permissions or hooks to be allowed without requiring approval
446- Changed indexed argument syntax from `$ARGUMENTS.0` to `$ARGUMENTS[0]` (bracket syntax)
447- [SDK] Added replay of `queued_command` attachment messages as `SDKUserMessageReplay` events when `replayUserMessages` is enabled
448- [VSCode] Enabled session forking and rewind functionality for all users
449
450## 2.1.18
451
452- Added customizable keyboard shortcuts. Configure keybindings per context, create chord sequences, and personalize your workflow. Run `/keybindings` to get started. Learn more at https://code.claude.com/docs/en/keybindings
453
454## 2.1.17
455
456- Fixed crashes on processors without AVX instruction support
457
458## 2.1.16
459
460- Added new task management system, including new capabilities like dependency tracking
461- [VSCode] Added native plugin management support
462- [VSCode] Added ability for OAuth users to browse and resume remote Claude sessions from the Sessions dialog
463- Fixed out-of-memory crashes when resuming sessions with heavy subagent usage
464- Fixed an issue where the "context remaining" warning was not hidden after running `/compact`
465- Fixed session titles on the resume screen not respecting the user's language setting
466- [IDE] Fixed a race condition on Windows where the Claude Code sidebar view container would not appear on start
467
468## 2.1.15
469
470- Added deprecation notification for npm installations - run `claude install` or see https://docs.anthropic.com/en/docs/claude-code/getting-started for more options
471- Improved UI rendering performance with React Compiler
472- Fixed the "Context left until auto-compact" warning not disappearing after running `/compact`
473- Fixed MCP stdio server timeout not killing child process, which could cause UI freezes
474
475## 2.1.14
476
477- Added history-based autocomplete in bash mode (`!`) - type a partial command and press Tab to complete from your bash command history
478- Added search to installed plugins list - type to filter by name or description
479- Added support for pinning plugins to specific git commit SHAs, allowing marketplace entries to install exact versions
480- Fixed a regression where the context window blocking limit was calculated too aggressively, blocking users at ~65% context usage instead of the intended ~98%
481- Fixed memory issues that could cause crashes when running parallel subagents
482- Fixed memory leak in long-running sessions where stream resources were not cleaned up after shell commands completed
483- Fixed `@` symbol incorrectly triggering file autocomplete suggestions in bash mode
484- Fixed `@`-mention menu folder click behavior to navigate into directories instead of selecting them
485- Fixed `/feedback` command generating invalid GitHub issue URLs when description is very long
486- Fixed `/context` command to show the same token count and percentage as the status line in verbose mode
487- Fixed an issue where `/config`, `/context`, `/model`, and `/todos` command overlays could close unexpectedly
488- Fixed slash command autocomplete selecting wrong command when typing similar commands (e.g., `/context` vs `/compact`)
489- Fixed inconsistent back navigation in plugin marketplace when only one marketplace is configured
490- Fixed iTerm2 progress bar not clearing properly on exit, preventing lingering indicators and bell sounds
491- Improved backspace to delete pasted text as a single token instead of one character at a time
492- [VSCode] Added `/usage` command to display current plan usage
493
494## 2.1.12
495
496- Fixed message rendering bug
497
498## 2.1.11
499
500- Fixed excessive MCP connection requests for HTTP/SSE transports
501
502## 2.1.10
503
504- Added new `Setup` hook event that can be triggered via `--init`, `--init-only`, or `--maintenance` CLI flags for repository setup and maintenance operations
505- Added keyboard shortcut 'c' to copy OAuth URL when browser doesn't open automatically during login
506- Fixed a crash when running bash commands containing heredocs with JavaScript template literals like `${index + 1}`
507- Improved startup to capture keystrokes typed before the REPL is fully ready
508- Improved file suggestions to show as removable attachments instead of inserting text when accepted
509- [VSCode] Added install count display to plugin listings
510- [VSCode] Added trust warning when installing plugins
511
512## 2.1.9
513
514- Added `auto:N` syntax for configuring the MCP tool search auto-enable threshold, where N is the context window percentage (0-100)
515- Added `plansDirectory` setting to customize where plan files are stored
516- Added external editor support (Ctrl+G) in AskUserQuestion "Other" input field
517- Added session URL attribution to commits and PRs created from web sessions
518- Added support for `PreToolUse` hooks to return `additionalContext` to the model
519- Added `${CLAUDE_SESSION_ID}` string substitution for skills to access the current session ID
520- Fixed long sessions with parallel tool calls failing with an API error about orphan tool_result blocks
521- Fixed MCP server reconnection hanging when cached connection promise never resolves
522- Fixed Ctrl+Z suspend not working in terminals using Kitty keyboard protocol (Ghostty, iTerm2, kitty, WezTerm)
523
524## 2.1.7
525
526- Added `showTurnDuration` setting to hide turn duration messages (e.g., "Cooked for 1m 6s")
527- Added ability to provide feedback when accepting permission prompts
528- Added inline display of agent's final response in task notifications, making it easier to see results without reading the full transcript file
529- Fixed security vulnerability where wildcard permission rules could match compound commands containing shell operators
530- Fixed false "file modified" errors on Windows when cloud sync tools, antivirus scanners, or Git touch file timestamps without changing content
531- Fixed orphaned tool_result errors when sibling tools fail during streaming execution
532- 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)
533- Fixed spinner briefly flashing when running local slash commands like `/model` or `/theme`
534- Fixed terminal title animation jitter by using fixed-width braille characters
535- Fixed plugins with git submodules not being fully initialized when installed
536- Fixed bash commands failing on Windows when temp directory paths contained characters like `t` or `n` that were misinterpreted as escape sequences
537- Improved typing responsiveness by reducing memory allocation overhead in terminal rendering
538- 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.
539- Changed OAuth and API Console URLs from console.anthropic.com to platform.claude.com
540- [VSCode] Fixed `claudeProcessWrapper` setting passing the wrapper path instead of the Claude binary path
541
542## 2.1.6
543
544- Added search functionality to `/config` command for quickly filtering settings
545- Added Updates section to `/doctor` showing auto-update channel and available npm versions (stable/latest)
546- Added date range filtering to `/stats` command - press `r` to cycle between Last 7 days, Last 30 days, and All time
547- Added automatic discovery of skills from nested `.claude/skills` directories when working with files in subdirectories
548- Added `context_window.used_percentage` and `context_window.remaining_percentage` fields to status line input for easier context window display
549- Added an error display when the editor fails during Ctrl+G
550- Fixed permission bypass via shell line continuation that could allow blocked commands to execute
551- Fixed false "File has been unexpectedly modified" errors when file watchers touch files without changing content
552- Fixed text styling (bold, colors) getting progressively misaligned in multi-line responses
553- Fixed the feedback panel closing unexpectedly when typing 'n' in the description field
554- Fixed rate limit warning appearing at low usage after weekly reset (now requires 70% usage)
555- Fixed rate limit options menu incorrectly auto-opening when resuming a previous session
556- Fixed numpad keys outputting escape sequences instead of characters in Kitty keyboard protocol terminals
557- Fixed Option+Return not inserting newlines in Kitty keyboard protocol terminals
558- Fixed corrupted config backup files accumulating in the home directory (now only one backup is created per config file)
559- Fixed `mcp list` and `mcp get` commands leaving orphaned MCP server processes
560- Fixed visual artifacts in ink2 mode when nodes become hidden via `display:none`
561- Improved the external CLAUDE.md imports approval dialog to show which files are being imported and from where
562- Improved the `/tasks` dialog to go directly to task details when there's only one background task running
563- Improved @ autocomplete with icons for different suggestion types and single-line formatting
564- Updated "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
565- Changed task notification display to cap at 3 lines with overflow summary when multiple background tasks complete simultaneously
566- Changed terminal title to "Claude Code" on startup for better window identification
567- Removed ability to @-mention MCP servers to enable/disable - use `/mcp enable <name>` instead
568- [VSCode] Fixed usage indicator not updating after manual compact
569
570## 2.1.5
571
572- Added `CLAUDE_CODE_TMPDIR` environment variable to override the temp directory used for internal temp files, useful for environments with custom temp directory requirements
573
574## 2.1.4
575
576- Added `CLAUDE_CODE_DISABLE_BACKGROUND_TASKS` environment variable to disable all background task functionality including auto-backgrounding and the Ctrl+B shortcut
577- Fixed "Help improve Claude" setting fetch to refresh OAuth and retry when it fails due to a stale OAuth token
578
579## 2.1.3
580
581- Merged slash commands and skills, simplifying the mental model with no change in behavior
582- Added release channel (`stable` or `latest`) toggle to `/config`
583- 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
584- Fixed plan files persisting across `/clear` commands, now ensuring a fresh plan file is used after clearing a conversation
585- Fixed false skill duplicate detection on filesystems with large inodes (e.g., ExFAT) by using 64-bit precision for inode values
586- Fixed mismatch between background task count in status bar and items shown in tasks dialog
587- Fixed sub-agents using the wrong model during conversation compaction
588- Fixed web search in sub-agents using incorrect model
589- Fixed trust dialog acceptance when running from the home directory not enabling trust-requiring features like hooks during the session
590- Improved terminal rendering stability by preventing uncontrolled writes from corrupting cursor state
591- Improved slash command suggestion readability by truncating long descriptions to 2 lines
592- Changed tool hook execution timeout from 60 seconds to 10 minutes
593- [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)
594
595## 2.1.2
596
597- Added source path metadata to images dragged onto the terminal, helping Claude understand where images originated
598- Added clickable hyperlinks for file paths in tool output in terminals that support OSC 8 (like iTerm)
599- Added support for Windows Package Manager (winget) installations with automatic detection and update instructions
600- Added Shift+Tab keyboard shortcut in plan mode to quickly select "auto-accept edits" option
601- Added `FORCE_AUTOUPDATE_PLUGINS` environment variable to allow plugin autoupdate even when the main auto-updater is disabled
602- Added `agent_type` to SessionStart hook input, populated if `--agent` is specified
603- Fixed a command injection vulnerability in bash command processing where malformed input could execute arbitrary commands
604- Fixed a memory leak where tree-sitter parse trees were not being freed, causing WASM memory to grow unbounded over long sessions
605- Fixed binary files (images, PDFs, etc.) being accidentally included in memory when using `@include` directives in CLAUDE.md files
606- Fixed updates incorrectly claiming another installation is in progress
607- Fixed crash when socket files exist in watched directories (defense-in-depth for EOPNOTSUPP errors)
608- Fixed remote session URL and teleport being broken when using `/tasks` command
609- Fixed MCP tool names being exposed in analytics events by sanitizing user-specific server configurations
610- Improved Option-as-Meta hint on macOS to show terminal-specific instructions for native CSIu terminals like iTerm2, Kitty, and WezTerm
611- Improved error message when pasting images over SSH to suggest using `scp` instead of the unhelpful clipboard shortcut hint
612- Improved permission explainer to not flag routine dev workflows (git fetch/rebase, npm install, tests, PRs) as medium risk
613- Changed large bash command outputs to be saved to disk instead of truncated, allowing Claude to read the full content
614- Changed large tool outputs to be persisted to disk instead of truncated, providing full output access via file references
615- Changed `/plugins` installed tab to unify plugins and MCPs with scope-based grouping
616- Deprecated Windows managed settings path `C:\ProgramData\ClaudeCode\managed-settings.json` - administrators should migrate to `C:\Program Files\ClaudeCode\managed-settings.json`
617- [SDK] Changed minimum zod peer dependency to ^4.0.0
618- [VSCode] Fixed usage display not updating after manual compact
619
620## 2.1.0
621
622- Added automatic skill hot-reload - skills created or modified in `~/.claude/skills` or `.claude/skills` are now immediately available without restarting the session
623- Added support for running skills and slash commands in a forked sub-agent context using `context: fork` in skill frontmatter
624- Added support for `agent` field in skills to specify agent type for execution
625- Added `language` setting to configure Claude's response language (e.g., language: "japanese")
626- Changed Shift+Enter to work out of the box in iTerm2, WezTerm, Ghostty, and Kitty without modifying terminal configs
627- Added `respectGitignore` support in `settings.json` for per-project control over @-mention file picker behavior
628- Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
629- Fixed security issue where sensitive data (OAuth tokens, API keys, passwords) could be exposed in debug logs
630- Fixed files and skills not being properly discovered when resuming sessions with `-c` or `--resume`
631- Fixed pasted content being lost when replaying prompts from history using up arrow or Ctrl+R search
632- Fixed Esc key with queued prompts to only move them to input without canceling the running task
633- Reduced permission prompts for complex bash commands
634- Fixed command search to prioritize exact and prefix matches on command names over fuzzy matches in descriptions
635- Fixed PreToolUse hooks to allow `updatedInput` when returning `ask` permission decision, enabling hooks to act as middleware while still requesting user consent
636- Fixed plugin path resolution for file-based marketplace sources
637- Fixed LSP tool being incorrectly enabled when no LSP servers were configured
638- Fixed background tasks failing with "git repository not found" error for repositories with dots in their names
639- Fixed Claude in Chrome support for WSL environments
640- Fixed Windows native installer silently failing when executable creation fails
641- Improved CLI help output to display options and subcommands in alphabetical order for easier navigation
642- Added wildcard pattern matching for Bash tool permissions using `*` at any position in rules (e.g., `Bash(npm *)`, `Bash(* install)`, `Bash(git * main)`)
643- Added unified Ctrl+B backgrounding for both bash commands and agents - pressing Ctrl+B now backgrounds all running foreground tasks simultaneously
644- Added support for MCP `list_changed` notifications, allowing MCP servers to dynamically update their available tools, prompts, and resources without requiring reconnection
645- Added `/teleport` and `/remote-env` slash commands for claude.ai subscribers, allowing them to resume and configure remote sessions
646- Added support for disabling specific agents using `Task(AgentName)` syntax in settings.json permissions or the `--disallowedTools` CLI flag
647- Added hooks support to agent frontmatter, allowing agents to define PreToolUse, PostToolUse, and Stop hooks scoped to the agent's lifecycle
648- Added hooks support for skill and slash command frontmatter
649- 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
650- Added `/plan` command shortcut to enable plan mode directly from the prompt
651- Added slash command autocomplete support when `/` appears anywhere in input, not just at the beginning
652- Added `--tools` flag support in interactive mode to restrict which built-in tools Claude can use during interactive sessions
653- Added `CLAUDE_CODE_FILE_READ_MAX_OUTPUT_TOKENS` environment variable to override the default file read token limit
654- Added support for `once: true` config for hooks
655- Added support for YAML-style lists in frontmatter `allowed-tools` field for cleaner skill declarations
656- Added support for prompt and agent hook types from plugins (previously only command hooks were supported)
657- Added Cmd+V support for image paste in iTerm2 (maps to Ctrl+V)
658- Added left/right arrow key navigation for cycling through tabs in dialogs
659- Added real-time thinking block display in Ctrl+O transcript mode
660- Added filepath to full output in background bash task details dialog
661- Added Skills as a separate category in the context visualization
662- Fixed OAuth token refresh not triggering when server reports token expired but local expiration check disagrees
663- Fixed session persistence getting stuck after transient server errors by recovering from 409 conflicts when the entry was actually stored
664- Fixed session resume failures caused by orphaned tool results during concurrent tool execution
665- Fixed a race condition where stale OAuth tokens could be read from the keychain cache during concurrent token refresh attempts
666- Fixed AWS Bedrock subagents not inheriting EU/APAC cross-region inference model configuration, causing 403 errors when IAM permissions are scoped to specific regions
667- Fixed API context overflow when background tasks produce large output by truncating to 30K chars with file path reference
668- Fixed a hang when reading FIFO files by skipping symlink resolution for special file types
669- Fixed terminal keyboard mode not being reset on exit in Ghostty, iTerm2, Kitty, and WezTerm
670- Fixed Alt+B and Alt+F (word navigation) not working in iTerm2, Ghostty, Kitty, and WezTerm
671- Fixed `${CLAUDE_PLUGIN_ROOT}` not being substituted in plugin `allowed-tools` frontmatter, which caused tools to incorrectly require approval
672- Fixed files created by the Write tool using hardcoded 0o600 permissions instead of respecting the system umask
673- Fixed commands with `$()` command substitution failing with parse errors
674- Fixed multi-line bash commands with backslash continuations being incorrectly split and flagged for permissions
675- 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)
676- Fixed slash commands passed as CLI arguments (e.g., `claude /context`) not being executed properly
677- Fixed pressing Enter after Tab-completing a slash command selecting a different command instead of submitting the completed one
678- Fixed slash command argument hint flickering and inconsistent display when typing commands with arguments
679- Fixed Claude sometimes redundantly invoking the Skill tool when running slash commands directly
680- Fixed skill token estimates in `/context` to accurately reflect frontmatter-only loading
681- Fixed subagents sometimes not inheriting the parent's model by default
682- Fixed model picker showing incorrect selection for Bedrock/Vertex users using `--model haiku`
683- Fixed duplicate Bash commands appearing in permission request option labels
684- Fixed noisy output when background tasks complete - now shows clean completion message instead of raw output
685- Fixed background task completion notifications to appear proactively with bullet point
686- Fixed forked slash commands showing "AbortError" instead of "Interrupted" message when cancelled
687- Fixed cursor disappearing after dismissing permission dialogs
688- Fixed `/hooks` menu selecting wrong hook type when scrolling to a different option
689- Fixed images in queued prompts showing as "[object Object]" when pressing Esc to cancel
690- Fixed images being silently dropped when queueing messages while backgrounding a task
691- Fixed large pasted images failing with "Image was too large" error
692- Fixed extra blank lines in multiline prompts containing CJK characters (Japanese, Chinese, Korean)
693- Fixed ultrathink keyword highlighting being applied to wrong characters when user prompt text wraps to multiple lines
694- Fixed collapsed "Reading X files…" indicator incorrectly switching to past tense when thinking blocks appear mid-stream
695- Fixed Bash read commands (like `ls` and `cat`) not being counted in collapsed read/search groups, causing groups to incorrectly show "Read 0 files"
696- Fixed spinner token counter to properly accumulate tokens from subagents during execution
697- Fixed memory leak in git diff parsing where sliced strings retained large parent strings
698- Fixed race condition where LSP tool could return "no server available" during startup
699- Fixed feedback submission hanging indefinitely when network requests timeout
700- Fixed search mode in plugin discovery and log selector views exiting when pressing up arrow
701- Fixed hook success message showing trailing colon when hook has no output
702- Multiple optimizations to improve startup performance
703- Improved terminal rendering performance when using native installer or Bun, especially for text with emoji, ANSI codes, and Unicode characters
704- Improved performance when reading Jupyter notebooks with many cells
705- Improved reliability for piped input like `cat refactor.md | claude`
706- Improved reliability for AskQuestion tool
707- Improved sed in-place edit commands to render as file edits with diff preview
708- Improved Claude to automatically continue when response is cut off due to output token limit, instead of showing an error message
709- Improved compaction reliability
710- Improved subagents (Task tool) to continue working after permission denial, allowing them to try alternative approaches
711- Improved skills to show progress while executing, displaying tool uses as they happen
712- Improved skills from `/skills/` directories to be visible in the slash command menu by default (opt-out with `user-invocable: false` in frontmatter)
713- Improved skill suggestions to prioritize recently and frequently used skills
714- Improved spinner feedback when waiting for the first response token
715- Improved token count display in spinner to include tokens from background agents
716- Improved incremental output for async agents to give the main thread more control and visibility
717- Improved permission prompt UX with Tab hint moved to footer, cleaner Yes/No input labels with contextual placeholders
718- Improved the Claude in Chrome notification with shortened help text and persistent display until dismissed
719- Improved macOS screenshot paste reliability with TIFF format support
720- Improved `/stats` output
721- Updated Atlassian MCP integration to use a more reliable default configuration (streamable HTTP)
722- Changed "Interrupted" message color from red to grey for a less alarming appearance
723- Removed permission prompt when entering plan mode - users can now enter plan mode without approval
724- Removed underline styling from image reference links
725- [SDK] Changed minimum zod peer dependency to ^4.0.0
726- [VSCode] Added currently selected model name to the context menu
727- [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")
728- [VSCode] Fixed paragraph breaks not rendering in markdown content
729- [VSCode] Fixed scrolling in the extension inadvertently scrolling the parent iframe
730- [Windows] Fixed issue with improper rendering
731
732## 2.0.76
733
734- Fixed issue with macOS code-sign warning when using Claude in Chrome integration
735
736## 2.0.75
737
738- Minor bugfixes
739
740## 2.0.74
741
742- Added LSP (Language Server Protocol) tool for code intelligence features like go-to-definition, find references, and hover documentation
743- Added `/terminal-setup` support for Kitty, Alacritty, Zed, and Warp terminals
744- Added ctrl+t shortcut in `/theme` to toggle syntax highlighting on/off
745- Added syntax highlighting info to theme picker
746- Added guidance for macOS users when Alt shortcuts fail due to terminal configuration
747- Fixed skill `allowed-tools` not being applied to tools invoked by the skill
748- Fixed Opus 4.5 tip incorrectly showing when user was already using Opus
749- Fixed a potential crash when syntax highlighting isn't initialized correctly
750- Fixed visual bug in `/plugins discover` where list selection indicator showed while search box was focused
751- Fixed macOS keyboard shortcuts to display 'opt' instead of 'alt'
752- Improved `/context` command visualization with grouped skills and agents by source, slash commands, and sorted token count
753- [Windows] Fixed issue with improper rendering
754- [VSCode] Added gift tag pictogram for year-end promotion message
755
756## 2.0.73
757
758- Added clickable `[Image #N]` links that open attached images in the default viewer
759- Added alt-y yank-pop to cycle through kill ring history after ctrl-y yank
760- Added search filtering to the plugin discover screen (type to filter by name, description, or marketplace)
761- Added support for custom session IDs when forking sessions with `--session-id` combined with `--resume` or `--continue` and `--fork-session`
762- Fixed slow input history cycling and race condition that could overwrite text after message submission
763- Improved `/theme` command to open theme picker directly
764- Improved theme picker UI
765- Improved search UX across resume session, permissions, and plugins screens with a unified SearchBox component
766- [VSCode] Added tab icon badges showing pending permissions (blue) and unread completions (orange)
767
768## 2.0.72
769
770- 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
771- Reduced terminal flickering
772- Added scannable QR code to mobile app tip for quick app downloads
773- Added loading indicator when resuming conversations for better feedback
774- Fixed `/context` command not respecting custom system prompts in non-interactive mode
775- Fixed order of consecutive Ctrl+K lines when pasting with Ctrl+Y
776- Improved @ mention file suggestion speed (~3× faster in git repositories)
777- Improved file suggestion performance in repos with `.ignore` or `.rgignore` files
778- Improved settings validation errors to be more prominent
779- Changed thinking toggle from Tab to Alt+T to avoid accidental triggers
780
781## 2.0.71
782
783- Added /config toggle to enable/disable prompt suggestions
784- Added `/settings` as an alias for the `/config` command
785- Fixed @ file reference suggestions incorrectly triggering when cursor is in the middle of a path
786- Fixed MCP servers from `.mcp.json` not loading when using `--dangerously-skip-permissions`
787- Fixed permission rules incorrectly rejecting valid bash commands containing shell glob patterns (e.g., `ls *.txt`, `for f in *.png`)
788- Bedrock: Environment variable `ANTHROPIC_BEDROCK_BASE_URL` is now respected for token counting and inference profile listing
789- New syntax highlighting engine for native build
790
791## 2.0.70
792
793- Added Enter key to accept and submit prompt suggestions immediately (tab still accepts for editing)
794- Added wildcard syntax `mcp__server__*` for MCP tool permissions to allow or deny all tools from a server
795- Added auto-update toggle for plugin marketplaces, allowing per-marketplace control over automatic updates
796- Added `current_usage` field to status line input, enabling accurate context window percentage calculations
797- Fixed input being cleared when processing queued commands while the user was typing
798- Fixed prompt suggestions replacing typed input when pressing Tab
799- Fixed diff view not updating when terminal is resized
800- Improved memory usage by 3x for large conversations
801- Improved resolution of stats screenshots copied to clipboard (Ctrl+S) for crisper images
802- Removed # shortcut for quick memory entry (tell Claude to edit your CLAUDE.md instead)
803- Fix thinking mode toggle in /config not persisting correctly
804- Improve UI for file creation permission dialog
805
806## 2.0.69
807
808- Minor bugfixes
809
810## 2.0.68
811
812- Fixed IME (Input Method Editor) support for languages like Chinese, Japanese, and Korean by correctly positioning the composition window at the cursor
813- Fixed a bug where disallowed MCP tools were visible to the model
814- Fixed an issue where steering messages could be lost while a subagent is working
815- Fixed Option+Arrow word navigation treating entire CJK (Chinese, Japanese, Korean) text sequences as a single word instead of navigating by word boundaries
816- Improved plan mode exit UX: show simplified yes/no dialog when exiting with empty or missing plan instead of throwing an error
817- Add support for enterprise managed settings. Contact your Anthropic account team to enable this feature.
818
819## 2.0.67
820
821- Thinking mode is now enabled by default for Opus 4.5
822- Thinking mode configuration has moved to /config
823- Added search functionality to `/permissions` command with `/` keyboard shortcut for filtering rules by tool name
824- Show reason why autoupdater is disabled in `/doctor`
825- Fixed false "Another process is currently updating Claude" error when running `claude update` while another instance is already on the latest version
826- Fixed MCP servers from `.mcp.json` being stuck in pending state when running in non-interactive mode (`-p` flag or piped input)
827- Fixed scroll position resetting after deleting a permission rule in `/permissions`
828- 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
829- Fixed `claude install --force` not bypassing stale lock files
830- Fixed consecutive @~/ file references in CLAUDE.md being incorrectly parsed due to markdown strikethrough interference
831- Windows: Fixed plugin MCP servers failing due to colons in log directory paths
832
833## 2.0.65
834
835- Added ability to switch models while writing a prompt using alt+p (linux, windows), option+p (macos).
836- Added context window information to status line input
837- Added `fileSuggestion` setting for custom `@` file search commands
838- Added `CLAUDE_CODE_SHELL` environment variable to override automatic shell detection (useful when login shell differs from actual working shell)
839- Fixed prompt not being saved to history when aborting a query with Escape
840- Fixed Read tool image handling to identify format from bytes instead of file extension
841
842## 2.0.64
843
844- Made auto-compacting instant
845- Agents and bash commands can run asynchronously and send messages to wake up the main agent
846- /stats now provides users with interesting CC stats, such as favorite model, usage graph, usage streak
847- Added named session support: use `/rename` to name sessions, `/resume <name>` in REPL or `claude --resume <name>` from the terminal to resume them
848- Added support for .claude/rules/`. See https://code.claude.com/docs/en/memory for details.
849- Added image dimension metadata when images are resized, enabling accurate coordinate mappings for large images
850- Fixed auto-loading .env when using native installer
851- Fixed `--system-prompt` being ignored when using `--continue` or `--resume` flags
852- Improved `/resume` screen with grouped forked sessions and keyboard shortcuts for preview (P) and rename (R)
853- VSCode: Added copy-to-clipboard button on code blocks and bash tool inputs
854- VSCode: Fixed extension not working on Windows ARM64 by falling back to x64 binary via emulation
855- Bedrock: Improve efficiency of token counting
856- Bedrock: Add support for `aws login` AWS Management Console credentials
857- Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
858
859## 2.0.62
860
861- Added "(Recommended)" indicator for multiple-choice questions, with the recommended option moved to the top of the list
862- Added `attribution` setting to customize commit and PR bylines (deprecates `includeCoAuthoredBy`)
863- Fixed duplicate slash commands appearing when ~/.claude is symlinked to a project directory
864- Fixed slash command selection not working when multiple commands share the same name
865- Fixed an issue where skill files inside symlinked skill directories could become circular symlinks
866- Fixed running versions getting removed because lock file incorrectly going stale
867- Fixed IDE diff tab not closing when rejecting file changes
868
869## 2.0.61
870
871- Reverted VSCode support for multiple terminal clients due to responsiveness issues.
872
873## 2.0.60
874
875- Added background agent support. Agents run in the background while you work
876- Added --disable-slash-commands CLI flag to disable all slash commands
877- Added model name to "Co-Authored-By" commit messages
878- Enabled "/mcp enable [server-name]" or "/mcp disable [server-name]" to quickly toggle all servers
879- Updated Fetch to skip summarization for pre-approved websites
880- VSCode: Added support for multiple terminal clients connecting to the IDE server simultaneously
881
882## 2.0.59
883
884- Added --agent CLI flag to override the agent setting for the current session
885- Added `agent` setting to configure main thread with a specific agent's system prompt, tool restrictions, and model
886- VS Code: Fixed .claude.json config file being read from incorrect location
887
888## 2.0.58
889
890- Pro users now have access to Opus 4.5 as part of their subscription!
891- Fixed timer duration showing "11m 60s" instead of "12m 0s"
892- Windows: Managed settings now prefer `C:\Program Files\ClaudeCode` if it exists. Support for `C:\ProgramData\ClaudeCode` will be removed in a future version.
893
894## 2.0.57
895
896- Added feedback input when rejecting plans, allowing users to tell Claude what to change
897- VSCode: Added streaming message support for real-time response display
898
899## 2.0.56
900
901- Added setting to enable/disable terminal progress bar (OSC 9;4)
902- 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.
903
904## 2.0.55
905
906- Fixed proxy DNS resolution being forced on by default. Now opt-in via `CLAUDE_CODE_PROXY_RESOLVES_HOSTS=true` environment variable
907- Fixed keyboard navigation becoming unresponsive when holding down arrow keys in memory location selector
908- Improved AskUserQuestion tool to auto-submit single-select questions on the last question, eliminating the extra review screen for simple question flows
909- Improved fuzzy matching for `@` file suggestions with faster, more accurate results
910
911## 2.0.54
912
913- Hooks: Enable PermissionRequest hooks to process 'always allow' suggestions and apply permission updates
914- Fix issue with excessive iTerm notifications
915
916## 2.0.52
917
918- Fixed duplicate message display when starting Claude with a command line argument
919- Fixed `/usage` command progress bars to fill up as usage increases (instead of showing remaining percentage)
920- Fixed image pasting not working on Linux systems running Wayland (now falls back to wl-paste when xclip is unavailable)
921- Permit some uses of `$!` in bash commands
922
923## 2.0.51
924
925- Added Opus 4.5! https://www.anthropic.com/news/claude-opus-4-5
926- Introducing Claude Code for Desktop: https://claude.com/download
927- 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
928- Pro users can now purchase extra usage for access to Opus 4.5 in Claude Code
929- Plan Mode now builds more precise plans and executes more thoroughly
930- Usage limit notifications now easier to understand
931- Switched `/usage` back to "% used"
932- Fixed handling of thinking errors
933- Fixed performance regression
934
935## 2.0.50
936
937- Fixed bug preventing calling MCP tools that have nested references in their input schemas
938- Silenced a noisy but harmless error during upgrades
939- Improved ultrathink text display
940- Improved clarity of 5-hour session limit warning message
941
942## 2.0.49
943
944- Added readline-style ctrl-y for pasting deleted text
945- Improved clarity of usage limit warning message
946- Fixed handling of subagent permissions
947
948## 2.0.47
949
950- Improved error messages and validation for `claude --teleport`
951- Improved error handling in `/usage`
952- Fixed race condition with history entry not getting logged at exit
953- Fixed Vertex AI configuration not being applied from `settings.json`
954
955## 2.0.46
956
957- Fixed image files being reported with incorrect media type when format cannot be detected from metadata
958
959## 2.0.45
960
961- Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
962- Added `PermissionRequest` hook to automatically approve or deny tool permission requests with custom logic
963- Send background tasks to Claude Code on the web by starting a message with `&`
964
965## 2.0.43
966
967- Added `permissionMode` field for custom agents
968- Added `tool_use_id` field to `PreToolUseHookInput` and `PostToolUseHookInput` types
969- Added skills frontmatter field to declare skills to auto-load for subagents
970- Added the `SubagentStart` hook event
971- Fixed nested `CLAUDE.md` files not loading when @-mentioning files
972- Fixed duplicate rendering of some messages in the UI
973- Fixed some visual flickers
974- Fixed NotebookEdit tool inserting cells at incorrect positions when cell IDs matched the pattern `cell-N`
975
976## 2.0.42
977
978- Added `agent_id` and `agent_transcript_path` fields to `SubagentStop` hooks.
979
980## 2.0.41
981
982- Added `model` parameter to prompt-based stop hooks, allowing users to specify a custom model for hook evaluation
983- Fixed slash commands from user settings being loaded twice, which could cause rendering issues
984- Fixed incorrect labeling of user settings vs project settings in command descriptions
985- Fixed crash when plugin command hooks timeout during execution
986- Fixed: Bedrock users no longer see duplicate Opus entries in the /model picker when using `--model haiku`
987- Fixed broken security documentation links in trust dialogs and onboarding
988- Fixed issue where pressing ESC to close the diff modal would also interrupt the model
989- ctrl-r history search landing on a slash command no longer cancels the search
990- SDK: Support custom timeouts for hooks
991- Allow more safe git commands to run without approval
992- Plugins: Added support for sharing and installing output styles
993- Teleporting a session from web will automatically set the upstream branch
994
995## 2.0.37
996
997- Fixed how idleness is computed for notifications
998- Hooks: Added matcher values for Notification hook events
999- Output Styles: Added `keep-coding-instructions` option to frontmatter
1000
1001## 2.0.36
1002
1003- Fixed: DISABLE_AUTOUPDATER environment variable now properly disables package manager update notifications
1004- Fixed queued messages being incorrectly executed as bash commands
1005- Fixed input being lost when typing while a queued message is processed
1006
1007## 2.0.35
1008
1009- Improve fuzzy search results when searching commands
1010- Improved VS Code extension to respect `chat.fontSize` and `chat.fontFamily` settings throughout the entire UI, and apply font changes immediately without requiring reload
1011- 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
1012- Migrated `ignorePatterns` from project config to deny permissions in the localSettings.
1013- Fixed menu navigation getting stuck on items with empty string or other falsy values (e.g., in the `/hooks` menu)
1014
1015## 2.0.34
1016
1017- VSCode Extension: Added setting to configure the initial permission mode for new conversations
1018- Improved file path suggestion performance with native Rust-based fuzzy finder
1019- Fixed infinite token refresh loop that caused MCP servers with OAuth (e.g., Slack) to hang during connection
1020- Fixed memory crash when reading or writing large files (especially base64-encoded images)
1021
1022## 2.0.33
1023
1024- Native binary installs now launch quicker.
1025- Fixed `claude doctor` incorrectly detecting Homebrew vs npm-global installations by properly resolving symlinks
1026- Fixed `claude mcp serve` exposing tools with incompatible outputSchemas
1027
1028## 2.0.32
1029
1030- Un-deprecate output styles based on community feedback
1031- Added `companyAnnouncements` setting for displaying announcements on startup
1032- Fixed hook progress messages not updating correctly during PostToolUse hook execution
1033
1034## 2.0.31
1035
1036- Windows: native installation uses shift+tab as shortcut for mode switching, instead of alt+m
1037- Vertex: add support for Web Search on supported models
1038- VSCode: Adding the respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
1039- Fixed a bug with subagents and MCP servers related to "Tool names must be unique" error
1040- Fixed issue causing `/compact` to fail with `prompt_too_long` by making it respect existing compact boundaries
1041- Fixed plugin uninstall not removing plugins
1042
1043## 2.0.30
1044
1045- Added helpful hint to run `security unlock-keychain` when encountering API key errors on macOS with locked keychain
1046- Added `allowUnsandboxedCommands` sandbox setting to disable the dangerouslyDisableSandbox escape hatch at policy level
1047- Added `disallowedTools` field to custom agent definitions for explicit tool blocking
1048- Added prompt-based stop hooks
1049- VSCode: Added respectGitIgnore configuration to include .gitignored files in file searches (defaults to true)
1050- Enabled SSE MCP servers on native build
1051- Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
1052- Removed support for custom ripgrep configuration, resolving an issue where Search returns no results and config discovery fails
1053- Fixed Explore agent creating unwanted .md investigation files during codebase exploration
1054- Fixed a bug where `/context` would sometimes fail with "max_tokens must be greater than thinking.budget_tokens" error message
1055- Fixed `--mcp-config` flag to correctly override file-based MCP configurations
1056- Fixed bug that saved session permissions to local settings
1057- Fixed MCP tools not being available to sub-agents
1058- Fixed hooks and plugins not executing when using --dangerously-skip-permissions flag
1059- Fixed delay when navigating through typeahead suggestions with arrow keys
1060- VSCode: Restored selection indicator in input footer showing current file or code selection status
1061
1062## 2.0.28
1063
1064- Plan mode: introduced new Plan subagent
1065- Subagents: claude can now choose to resume subagents
1066- Subagents: claude can dynamically choose the model used by its subagents
1067- SDK: added --max-budget-usd flag
1068- Discovery of custom slash commands, subagents, and output styles no longer respects .gitignore
1069- Stop `/terminal-setup` from adding backslash to `Shift + Enter` in VS Code
1070- Add branch and tag support for git-based plugins and marketplaces using fragment syntax (e.g., `owner/repo#branch`)
1071- Fixed a bug where macOS permission prompts would show up upon initial launch when launching from home directory
1072- Various other bug fixes
1073
1074## 2.0.27
1075
1076- New UI for permission prompts
1077- Added current branch filtering and search to session resume screen for easier navigation
1078- Fixed directory @-mention causing "No assistant message found" error
1079- VSCode Extension: Add config setting to include .gitignored files in file searches
1080- VSCode Extension: Bug fixes for unrelated 'Warmup' conversations, and configuration/settings occasionally being reset to defaults
1081
1082## 2.0.25
1083
1084- 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
1085
1086## 2.0.24
1087
1088- Fixed a bug where project-level skills were not loading when --setting-sources 'project' was specified
1089- Claude Code Web: Support for Web -> CLI teleport
1090- Sandbox: Releasing a sandbox mode for the BashTool on Linux & Mac
1091- Bedrock: Display awsAuthRefresh output when auth is required
1092
1093## 2.0.22
1094
1095- Fixed content layout shift when scrolling through slash commands
1096- IDE: Add toggle to enable/disable thinking.
1097- Fix bug causing duplicate permission prompts with parallel tool calls
1098- Add support for enterprise managed MCP allowlist and denylist
1099
1100## 2.0.21
1101
1102- Support MCP `structuredContent` field in tool responses
1103- Added an interactive question tool
1104- Claude will now ask you questions more often in plan mode
1105- Added Haiku 4.5 as a model option for Pro users
1106- Fixed an issue where queued commands don't have access to previous messages' output
1107
1108## 2.0.20
1109
1110- Added support for Claude Skills
1111
1112## 2.0.19
1113
1114- Auto-background long-running bash commands instead of killing them. Customize with BASH_DEFAULT_TIMEOUT_MS
1115- Fixed a bug where Haiku was unnecessarily called in print mode
1116
1117## 2.0.17
1118
1119- Added Haiku 4.5 to model selector!
1120- Haiku 4.5 automatically uses Sonnet in plan mode, and Haiku for execution (i.e. SonnetPlan by default)
1121- 3P (Bedrock and Vertex) are not automatically upgraded yet. Manual upgrading can be done through setting `ANTHROPIC_DEFAULT_HAIKU_MODEL`
1122- Introducing the Explore subagent. Powered by Haiku it'll search through your codebase efficiently to save context!
1123- OTEL: support HTTP_PROXY and HTTPS_PROXY
1124- `CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC` now disables release notes fetching
1125
1126## 2.0.15
1127
1128- Fixed bug with resuming where previously created files needed to be read again before writing
1129- Fixed bug with `-p` mode where @-mentioned files needed to be read again before writing
1130
1131## 2.0.14
1132
1133- Fix @-mentioning MCP servers to toggle them on/off
1134- Improve permission checks for bash with inline env vars
1135- Fix ultrathink + thinking toggle
1136- Reduce unnecessary logins
1137- Document --system-prompt
1138- Several improvements to rendering
1139- Plugins UI polish
1140
1141## 2.0.13
1142
1143- Fixed `/plugin` not working on native build
1144
1145## 2.0.12
1146
1147- **Plugin System Released**: Extend Claude Code with custom commands, agents, hooks, and MCP servers from marketplaces
1148- `/plugin install`, `/plugin enable/disable`, `/plugin marketplace` commands for plugin management
1149- Repository-level plugin configuration via `extraKnownMarketplaces` for team collaboration
1150- `/plugin validate` command for validating plugin structure and configuration
1151- Plugin announcement blog post at https://www.anthropic.com/news/claude-code-plugins
1152- Plugin documentation available at https://code.claude.com/docs/en/plugins
1153- Comprehensive error messages and diagnostics via `/doctor` command
1154- Avoid flickering in `/model` selector
1155- Improvements to `/help`
1156- Avoid mentioning hooks in `/resume` summaries
1157- Changes to the "verbose" setting in `/config` now persist across sessions
1158
1159## 2.0.11
1160
1161- Reduced system prompt size by 1.4k tokens
1162- IDE: Fixed keyboard shortcuts and focus issues for smoother interaction
1163- Fixed Opus fallback rate limit errors appearing incorrectly
1164- Fixed /add-dir command selecting wrong default tab
1165
1166## 2.0.10
1167
1168- Rewrote terminal renderer for buttery smooth UI
1169- Enable/disable MCP servers by @mentioning, or in /mcp
1170- Added tab completion for shell commands in bash mode
1171- PreToolUse hooks can now modify tool inputs
1172- Press Ctrl-G to edit your prompt in your system's configured text editor
1173- Fixes for bash permission checks with environment variables in the command
1174
1175## 2.0.9
1176
1177- Fix regression where bash backgrounding stopped working
1178
1179## 2.0.8
1180
1181- Update Bedrock default Sonnet model to `global.anthropic.claude-sonnet-4-5-20250929-v1:0`
1182- IDE: Add drag-and-drop support for files and folders in chat
1183- /context: Fix counting for thinking blocks
1184- Improve message rendering for users with light themes on dark terminals
1185- Remove deprecated .claude.json allowedTools, ignorePatterns, env, and todoFeatureEnabled config options (instead, configure these in your settings.json)
1186
1187## 2.0.5
1188
1189- IDE: Fix IME unintended message submission with Enter and Tab
1190- IDE: Add "Open in Terminal" link in login screen
1191- Fix unhandled OAuth expiration 401 API errors
1192- SDK: Added SDKUserMessageReplay.isReplay to prevent duplicate messages
1193
1194## 2.0.1
1195
1196- Skip Sonnet 4.5 default model setting change for Bedrock and Vertex
1197- Various bug fixes and presentation improvements
1198
1199## 2.0.0
1200
1201- New native VS Code extension
1202- Fresh coat of paint throughout the whole app
1203- /rewind a conversation to undo code changes
1204- /usage command to see plan limits
1205- Tab to toggle thinking (sticky across sessions)
1206- Ctrl-R to search history
1207- Unshipped claude config command
1208- Hooks: Reduced PostToolUse 'tool_use' ids were found without 'tool_result' blocks errors
1209- SDK: The Claude Code SDK is now the Claude Agent SDK
1210- Add subagents dynamically with `--agents` flag
1211
1212## 1.0.126
1213
1214- Enable /context command for Bedrock and Vertex
1215- Add mTLS support for HTTP-based OpenTelemetry exporters
1216
1217## 1.0.124
1218
1219- Set `CLAUDE_BASH_NO_LOGIN` environment variable to 1 or true to to skip login shell for BashTool
1220- Fix Bedrock and Vertex environment variables evaluating all strings as truthy
1221- No longer inform Claude of the list of allowed tools when permission is denied
1222- Fixed security vulnerability in Bash tool permission checks
1223- Improved VSCode extension performance for large files
1224
1225## 1.0.123
1226
1227- Bash permission rules now support output redirections when matching (e.g., `Bash(python:*)` matches `python script.py > output.txt`)
1228- Fixed thinking mode triggering on negation phrases like "don't think"
1229- Fixed rendering performance degradation during token streaming
1230- Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
1231- Enhanced BashTool environment snapshot logging
1232- Fixed a bug where resuming a conversation in headless mode would sometimes enable thinking unnecessarily
1233- Migrated --debug logging to a file, to enable easy tailing & filtering
1234
1235## 1.0.120
1236
1237- Fix input lag during typing, especially noticeable with large prompts
1238- Improved VSCode extension command registry and sessions dialog user experience
1239- Enhanced sessions dialog responsiveness and visual feedback
1240- Fixed IDE compatibility issue by removing worktree support check
1241- Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
1242
1243## 1.0.119
1244
1245- Fix Windows issue where process visually freezes on entering interactive mode
1246- Support dynamic headers for MCP servers via headersHelper configuration
1247- Fix thinking mode not working in headless sessions
1248- Fix slash commands now properly update allowed tools instead of replacing them
1249
1250## 1.0.117
1251
1252- Add Ctrl-R history search to recall previous commands like bash/zsh
1253- Fix input lag while typing, especially on Windows
1254- Add sed command to auto-allowed commands in acceptEdits mode
1255- Fix Windows PATH comparison to be case-insensitive for drive letters
1256- Add permissions management hint to /add-dir output
1257
1258## 1.0.115
1259
1260- Improve thinking mode display with enhanced visual effects
1261- Type /t to temporarily disable thinking mode in your prompt
1262- Improve path validation for glob and grep tools
1263- Show condensed output for post-tool hooks to reduce visual clutter
1264- Fix visual feedback when loading state completes
1265- Improve UI consistency for permission request dialogs
1266
1267## 1.0.113
1268
1269- Deprecated piped input in interactive mode
1270- Move Ctrl+R keybinding for toggling transcript to Ctrl+O
1271
1272## 1.0.112
1273
1274- Transcript mode (Ctrl+R): Added the model used to generate each assistant message
1275- Addressed issue where some Claude Max users were incorrectly recognized as Claude Pro users
1276- Hooks: Added systemMessage support for SessionEnd hooks
1277- Added `spinnerTipsEnabled` setting to disable spinner tips
1278- IDE: Various improvements and bug fixes
1279
1280## 1.0.111
1281
1282- /model now validates provided model names
1283- Fixed Bash tool crashes caused by malformed shell syntax parsing
1284
1285## 1.0.110
1286
1287- /terminal-setup command now supports WezTerm
1288- MCP: OAuth tokens now proactively refresh before expiration
1289- Fixed reliability issues with background Bash processes
1290
1291## 1.0.109
1292
1293- SDK: Added partial message streaming support via `--include-partial-messages` CLI flag
1294
1295## 1.0.106
1296
1297- Windows: Fixed path permission matching to consistently use POSIX format (e.g., `Read(//c/Users/...)`)
1298
1299## 1.0.97
1300
1301- Settings: /doctor now validates permission rule syntax and suggests corrections
1302
1303## 1.0.94
1304
1305- Vertex: add support for global endpoints for supported models
1306- /memory command now allows direct editing of all imported memory files
1307- SDK: Add custom tools as callbacks
1308- Added /todos command to list current todo items
1309
1310## 1.0.93
1311
1312- Windows: Add alt + v shortcut for pasting images from clipboard
1313- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
1314
1315## 1.0.90
1316
1317- Settings file changes take effect immediately - no restart required
1318
1319## 1.0.88
1320
1321- Fixed issue causing "OAuth authentication is currently not supported"
1322- Status line input now includes `exceeds_200k_tokens`
1323- Fixed incorrect usage tracking in /cost.
1324- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
1325- Bedrock: Updated default Sonnet model to Sonnet 4
1326
1327## 1.0.86
1328
1329- Added /context to help users self-serve debug context issues
1330- SDK: Added UUID support for all SDK messages
1331- SDK: Added `--replay-user-messages` to replay user messages back to stdout
1332
1333## 1.0.85
1334
1335- Status line input now includes session cost info
1336- Hooks: Introduced SessionEnd hook
1337
1338## 1.0.84
1339
1340- Fix tool_use/tool_result id mismatch error when network is unstable
1341- Fix Claude sometimes ignoring real-time steering when wrapping up a task
1342- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
1343- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
1344
1345## 1.0.83
1346
1347- @-mention: Support files with spaces in path
1348- New shimmering spinner
1349
1350## 1.0.82
1351
1352- SDK: Add request cancellation support
1353- SDK: New additionalDirectories option to search custom paths, improved slash command processing
1354- Settings: Validation prevents invalid fields in .claude/settings.json files
1355- MCP: Improve tool name consistency
1356- Bash: Fix crash when Claude tries to automatically read large files
1357
1358## 1.0.81
1359
1360- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
1361- Agents: Fix custom agent loading when agent files are unparsable
1362
1363## 1.0.80
1364
1365- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
1366
1367## 1.0.77
1368
1369- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
1370- SDK: Add session support and permission denial tracking
1371- Fix token limit errors in conversation summarization
1372- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
1373
1374## 1.0.73
1375
1376- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
1377- MCP: Press Esc to cancel OAuth authentication flows
1378- Bash: Improved command validation and reduced false security warnings
1379- UI: Enhanced spinner animations and status line visual hierarchy
1380- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
1381
1382## 1.0.72
1383
1384- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
1385
1386## 1.0.71
1387
1388- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
1389- Customizable status line: add your terminal prompt to Claude Code with /statusline
1390
1391## 1.0.70
1392
1393- Performance: Optimized message rendering for better performance with large contexts
1394- Windows: Fixed native file search, ripgrep, and subagent functionality
1395- Added support for @-mentions in slash command arguments
1396
1397## 1.0.69
1398
1399- Upgraded Opus to version 4.1
1400
1401## 1.0.68
1402
1403- Fix incorrect model names being used for certain commands like `/pr-comments`
1404- 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.
1405- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
1406- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
1407- SDK: Added canUseTool callback support for tool confirmation
1408- Added `disableAllHooks` setting
1409- Improved file suggestions performance in large repos
1410
1411## 1.0.65
1412
1413- IDE: Fixed connection stability issues and error handling for diagnostics
1414- Windows: Fixed shell environment setup for users without .bashrc files
1415
1416## 1.0.64
1417
1418- Agents: Added model customization support - you can now specify which model an agent should use
1419- Agents: Fixed unintended access to the recursive agent tool
1420- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
1421- SDK: Fixed user input tracking across multi-turn conversations
1422- Added hidden files to file search and @-mention suggestions
1423
1424## 1.0.63
1425
1426- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
1427
1428## 1.0.62
1429
1430- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
1431- Hooks: Added SessionStart hook for new session initialization
1432- /add-dir command now supports typeahead for directory paths
1433- Improved network connectivity check reliability
1434
1435## 1.0.61
1436
1437- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
1438- Settings: Added `--settings` flag to load settings from a JSON file
1439- Settings: Fixed resolution of settings files paths that are symlinks
1440- OTEL: Fixed reporting of wrong organization after authentication changes
1441- Slash commands: Fixed permissions checking for allowed-tools with Bash
1442- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
1443- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
1444- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
1445
1446## 1.0.60
1447
1448- You can now create custom subagents for specialized tasks! Run /agents to get started
1449
1450## 1.0.59
1451
1452- SDK: Added tool confirmation support with canUseTool callback
1453- SDK: Allow specifying env for spawned process
1454- Hooks: Exposed PermissionDecision to hooks (including "ask")
1455- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
1456- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
1457
1458## 1.0.58
1459
1460- Added support for reading PDFs
1461- MCP: Improved server health status display in 'claude mcp list'
1462- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
1463
1464## 1.0.57
1465
1466- Added support for specifying a model in slash commands
1467- Improved permission messages to help Claude understand allowed tools
1468- Fix: Remove trailing newlines from bash output in terminal wrapping
1469
1470## 1.0.56
1471
1472- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
1473- Fixes for WSL IDE detection
1474- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
1475
1476## 1.0.55
1477
1478- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
1479- Windows: fixed Ctrl+Z crash
1480- SDK: Added ability to capture error logging
1481- Add --system-prompt-file option to override system prompt in print mode
1482
1483## 1.0.54
1484
1485- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
1486- Custom slash commands: Added argument-hint to frontmatter
1487- Windows: OAuth uses port 45454 and properly constructs browser URL
1488- Windows: mode switching now uses alt + m, and plan mode renders properly
1489- Shell: Switch to in-memory shell snapshot to fix file-related errors
1490
1491## 1.0.53
1492
1493- Updated @-mention file truncation from 100 lines to 2000 lines
1494- 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).
1495
1496## 1.0.52
1497
1498- Added support for MCP server instructions
1499
1500## 1.0.51
1501
1502- Added support for native Windows (requires Git for Windows)
1503- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
1504- Settings: /doctor can now help you identify and fix invalid setting files
1505- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
1506- Increased auto-compact warning threshold from 60% to 80%
1507- Fixed an issue with handling user directories with spaces for shell snapshots
1508- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
1509- Custom slash commands: Fixed user-level commands in subdirectories
1510- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
1511
1512## 1.0.48
1513
1514- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
1515- Added progress messages to Bash tool based on the last 5 lines of command output
1516- Added expanding variables support for MCP server configuration
1517- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
1518- Improved IDE extension path handling when Claude Code runs in WSL
1519- Hooks: Added a PreCompact hook
1520- Vim mode: Added c, f/F, t/T
1521
1522## 1.0.45
1523
1524- Redesigned Search (Grep) tool with new tool input parameters and features
1525- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
1526- Fixed config file corruption issue by enforcing atomic writes
1527- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
1528- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
1529- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
1530
1531## 1.0.44
1532
1533- New /export command lets you quickly export a conversation for sharing
1534- MCP: resource_link tool results are now supported
1535- MCP: tool annotations and tool titles now display in /mcp view
1536- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
1537
1538## 1.0.43
1539
1540- Fixed a bug where the theme selector was saving excessively
1541- Hooks: Added EPIPE system error handling
1542
1543## 1.0.42
1544
1545- Added tilde (`~`) expansion support to `/add-dir` command
1546
1547## 1.0.41
1548
1549- Hooks: Split Stop hook triggering into Stop and SubagentStop
1550- Hooks: Enabled optional timeout configuration for each command
1551- Hooks: Added "hook_event_name" to hook input
1552- Fixed a bug where MCP tools would display twice in tool list
1553- New tool parameters JSON for Bash tool in `tool_decision` event
1554
1555## 1.0.40
1556
1557- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
1558
1559## 1.0.39
1560
1561- New Active Time metric in OpenTelemetry logging
1562
1563## 1.0.38
1564
1565- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
1566
1567## 1.0.37
1568
1569- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
1570
1571## 1.0.36
1572
1573- Web search now takes today's date into context
1574- Fixed a bug where stdio MCP servers were not terminating properly on exit
1575
1576## 1.0.35
1577
1578- Added support for MCP OAuth Authorization Server discovery
1579
1580## 1.0.34
1581
1582- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
1583
1584## 1.0.33
1585
1586- Improved logging functionality with session ID support
1587- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
1588- Improvements to plan mode
1589
1590## 1.0.32
1591
1592- Updated loopback config for litellm
1593- Added forceLoginMethod setting to bypass login selection screen
1594
1595## 1.0.31
1596
1597- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
1598
1599## 1.0.30
1600
1601- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
1602- Improved file path autocomplete with filename matching
1603- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
1604- Enhanced jq regex support for complex filters with pipes and select
1605
1606## 1.0.29
1607
1608- Improved CJK character support in cursor navigation and rendering
1609
1610## 1.0.28
1611
1612- Slash commands: Fix selector display during history navigation
1613- Resizes images before upload to prevent API size limit errors
1614- Added XDG_CONFIG_HOME support to configuration directory
1615- Performance optimizations for memory usage
1616- New attributes (terminal.type, language) in OpenTelemetry logging
1617
1618## 1.0.27
1619
1620- Streamable HTTP MCP servers are now supported
1621- Remote MCP servers (SSE and HTTP) now support OAuth
1622- MCP resources can now be @-mentioned
1623- /resume slash command to switch conversations within Claude Code
1624
1625## 1.0.25
1626
1627- Slash commands: moved "project" and "user" prefixes to descriptions
1628- Slash commands: improved reliability for command discovery
1629- Improved support for Ghostty
1630- Improved web search reliability
1631
1632## 1.0.24
1633
1634- Improved /mcp output
1635- Fixed a bug where settings arrays got overwritten instead of merged
1636
1637## 1.0.23
1638
1639- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
1640- Released Python SDK: pip install claude-code-sdk to get started
1641
1642## 1.0.22
1643
1644- SDK: Renamed `total_cost` to `total_cost_usd`
1645
1646## 1.0.21
1647
1648- Improved editing of files with tab-based indentation
1649- Fix for tool_use without matching tool_result errors
1650- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
1651
1652## 1.0.18
1653
1654- Added --add-dir CLI argument for specifying additional working directories
1655- Added streaming input support without require -p flag
1656- Improved startup performance and session storage performance
1657- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
1658- Added detailed MCP server tools display (/mcp)
1659- MCP authentication and permission improvements
1660- Added auto-reconnection for MCP SSE connections on disconnect
1661- Fixed issue where pasted content was lost when dialogs appeared
1662
1663## 1.0.17
1664
1665- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
1666- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
1667- MCP server list UI improvements
1668- Update Claude Code process title to display "claude" instead of "node"
1669
1670## 1.0.11
1671
1672- Claude Code can now also be used with a Claude Pro subscription
1673- Added /upgrade for smoother switching to Claude Max plans
1674- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
1675- Improved shell configuration error handling
1676- Improved todo list handling during compaction
1677
1678## 1.0.10
1679
1680- Added markdown table support
1681- Improved streaming performance
1682
1683## 1.0.8
1684
1685- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
1686- Increased default otel interval from 1s -> 5s
1687- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
1688- Fixed a regression where search tools unnecessarily asked for permissions
1689- Added support for triggering thinking non-English languages
1690- Improved compacting UI
1691
1692## 1.0.7
1693
1694- Renamed /allowed-tools -> /permissions
1695- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
1696- Deprecated claude config commands in favor of editing settings.json
1697- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
1698- Improved error handling for /install-github-app
1699- Bugfixes, UI polish, and tool reliability improvements
1700
1701## 1.0.6
1702
1703- Improved edit reliability for tab-indented files
1704- Respect CLAUDE_CONFIG_DIR everywhere
1705- Reduced unnecessary tool permission prompts
1706- Added support for symlinks in @file typeahead
1707- Bugfixes, UI polish, and tool reliability improvements
1708
1709## 1.0.4
1710
1711- Fixed a bug where MCP tool errors weren't being parsed correctly
1712
1713## 1.0.1
1714
1715- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
1716- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
1717- Updated documentation links and OAuth process descriptions
1718
1719## 1.0.0
1720
1721- Claude Code is now generally available
1722- Introducing Sonnet 4 and Opus 4 models
1723
1724## 0.2.125
1725
1726- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
1727- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
1728
1729## 0.2.117
1730
1731- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
1732- Introduced settings.cleanupPeriodDays
1733- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
1734- Introduced --debug mode
1735
1736## 0.2.108
1737
1738- You can now send messages to Claude while it works to steer Claude in real-time
1739- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
1740- Fixed a bug where thinking was not working in -p mode
1741- Fixed a regression in /cost reporting
1742- Deprecated MCP wizard interface in favor of other MCP commands
1743- Lots of other bugfixes and improvements
1744
1745## 0.2.107
1746
1747- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
1748
1749## 0.2.106
1750
1751- MCP SSE server configs can now specify custom headers
1752- Fixed a bug where MCP permission prompt didn't always show correctly
1753
1754## 0.2.105
1755
1756- Claude can now search the web
1757- Moved system & account status to /status
1758- Added word movement keybindings for Vim
1759- Improved latency for startup, todo tool, and file edits
1760
1761## 0.2.102
1762
1763- Improved thinking triggering reliability
1764- Improved @mention reliability for images and folders
1765- You can now paste multiple large chunks into one prompt
1766
1767## 0.2.100
1768
1769- Fixed a crash caused by a stack overflow error
1770- Made db storage optional; missing db support disables --continue and --resume
1771
1772## 0.2.98
1773
1774- Fixed an issue where auto-compact was running twice
1775
1776## 0.2.96
1777
1778- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
1779
1780## 0.2.93
1781
1782- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
1783- Claude now has access to a Todo list that helps it stay on track and be more organized
1784
1785## 0.2.82
1786
1787- Added support for --disallowedTools
1788- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
1789
1790## 0.2.75
1791
1792- Hit Enter to queue up additional messages while Claude is working
1793- Drag in or copy/paste image files directly into the prompt
1794- @-mention files to directly add them to context
1795- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
1796- Improved performance for filename auto-complete
1797
1798## 0.2.74
1799
1800- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
1801- Task tool can now perform writes and run bash commands
1802
1803## 0.2.72
1804
1805- Updated spinner to indicate tokens loaded and tool usage
1806
1807## 0.2.70
1808
1809- Network commands like curl are now available for Claude to use
1810- Claude can now run multiple web queries in parallel
1811- Pressing ESC once immediately interrupts Claude in Auto-accept mode
1812
1813## 0.2.69
1814
1815- Fixed UI glitches with improved Select component behavior
1816- Enhanced terminal output display with better text truncation logic
1817
1818## 0.2.67
1819
1820- Shared project permission rules can be saved in .claude/settings.json
1821
1822## 0.2.66
1823
1824- Print mode (-p) now supports streaming output via --output-format=stream-json
1825- Fixed issue where pasting could trigger memory or bash mode unexpectedly
1826
1827## 0.2.63
1828
1829- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
1830
1831## 0.2.61
1832
1833- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
1834- Enhanced image detection for more reliable clipboard paste functionality
1835- Fixed an issue where ESC key could crash the conversation history selector
1836
1837## 0.2.59
1838
1839- Copy+paste images directly into your prompt
1840- Improved progress indicators for bash and fetch tools
1841- Bugfixes for non-interactive mode (-p)
1842
1843## 0.2.54
1844
1845- Quickly add to Memory by starting your message with '#'
1846- Press ctrl+r to see full output for long tool results
1847- Added support for MCP SSE transport
1848
1849## 0.2.53
1850
1851- New web fetch tool lets Claude view URLs that you paste in
1852- Fixed a bug with JPEG detection
1853
1854## 0.2.50
1855
1856- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
1857
1858## 0.2.49
1859
1860- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
1861
1862## 0.2.47
1863
1864- Press Tab to auto-complete file and folder names
1865- Press Shift + Tab to toggle auto-accept for file edits
1866- Automatic conversation compaction for infinite conversation length (toggle with /config)
1867
1868## 0.2.44
1869
1870- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
1871
1872## 0.2.41
1873
1874- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
1875- MCP server startup no longer blocks the app from starting up
1876
1877## 0.2.37
1878
1879- New /release-notes command lets you view release notes at any time
1880- `claude config add/remove` commands now accept multiple values separated by commas or spaces
1881
1882## 0.2.36
1883
1884- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
1885- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
1886
1887## 0.2.34
1888
1889- Vim bindings for text input - enable with /vim or /config
1890
1891## 0.2.32
1892
1893- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
1894- Fix for some PersistentShell issues
1895
1896## 0.2.31
1897
1898- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
1899- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
1900
1901## 0.2.30
1902
1903- Added ANSI color theme for better terminal compatibility
1904- Fixed issue where slash command arguments weren't being sent properly
1905- (Mac-only) API keys are now stored in macOS Keychain
1906
1907## 0.2.26
1908
1909- New /approved-tools command for managing tool permissions
1910- Word-level diff display for improved code readability
1911- Fuzzy matching for slash commands
1912
1913## 0.2.21
1914
1915- Fuzzy matching for /commands