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 eb0e434 · on Aug 22, 2025 History
599 lines · 21.0 KBPreviewCodeRaw
1# Changelog
2
3## 1.0.88
4
5- Fixed issue causing "OAuth authentication is currently not supported"
6- Status line input now includes `exceeds_200k_tokens`
7- Fixed incorrect usage tracking in /cost.
8- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
9- Bedrock: Updated default Sonnet model to Sonnet 4
10
11## 1.0.86
12
13- Added /context to help users self-serve debug context issues
14- SDK: Added UUID support for all SDK messages
15- SDK: Added `--replay-user-messages` to replay user messages back to stdout
16
17## 1.0.85
18
19- Status line input now includes session cost info
20- Hooks: Introduced SessionEnd hook
21
22## 1.0.84
23
24- Fix tool_use/tool_result id mismatch error when network is unstable
25- Fix Claude sometimes ignoring real-time steering when wrapping up a task
26- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
27- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
28
29## 1.0.83
30
31- @-mention: Support files with spaces in path
32- New shimmering spinner
33
34## 1.0.82
35
36- SDK: Add request cancellation support
37- SDK: New additionalDirectories option to search custom paths, improved slash command processing
38- Settings: Validation prevents invalid fields in .claude/settings.json files
39- MCP: Improve tool name consistency
40- Bash: Fix crash when Claude tries to automatically read large files
41
42## 1.0.81
43
44- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.anthropic.com/en/docs/claude-code/output-styles
45- Agents: Fix custom agent loading when agent files are unparsable
46
47## 1.0.80
48
49- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
50
51## 1.0.77
52
53- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
54- SDK: Add session support and permission denial tracking
55- Fix token limit errors in conversation summarization
56- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
57
58## 1.0.73
59
60- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
61- MCP: Press Esc to cancel OAuth authentication flows
62- Bash: Improved command validation and reduced false security warnings
63- UI: Enhanced spinner animations and status line visual hierarchy
64- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
65
66## 1.0.72
67
68- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
69
70## 1.0.71
71
72- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
73- Customizable status line: add your terminal prompt to Claude Code with /statusline
74
75## 1.0.70
76
77- Performance: Optimized message rendering for better performance with large contexts
78- Windows: Fixed native file search, ripgrep, and subagent functionality
79- Added support for @-mentions in slash command arguments
80
81## 1.0.69
82
83- Upgraded Opus to version 4.1
84
85## 1.0.68
86
87- Fix incorrect model names being used for certain commands like `/pr-comments`
88- 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.
89- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
90- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
91- SDK: Added canUseTool callback support for tool confirmation
92- Added `disableAllHooks` setting
93- Improved file suggestions performance in large repos
94
95## 1.0.65
96
97- IDE: Fixed connection stability issues and error handling for diagnostics
98- Windows: Fixed shell environment setup for users without .bashrc files
99
100## 1.0.64
101
102- Agents: Added model customization support - you can now specify which model an agent should use
103- Agents: Fixed unintended access to the recursive agent tool
104- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
105- SDK: Fixed user input tracking across multi-turn conversations
106- Added hidden files to file search and @-mention suggestions
107
108## 1.0.63
109
110- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
111
112## 1.0.62
113
114- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
115- Hooks: Added SessionStart hook for new session initialization
116- /add-dir command now supports typeahead for directory paths
117- Improved network connectivity check reliability
118
119## 1.0.61
120
121- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
122- Settings: Added `--settings` flag to load settings from a JSON file
123- Settings: Fixed resolution of settings files paths that are symlinks
124- OTEL: Fixed reporting of wrong organization after authentication changes
125- Slash commands: Fixed permissions checking for allowed-tools with Bash
126- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
127- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
128- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
129
130## 1.0.60
131
132- You can now create custom subagents for specialized tasks! Run /agents to get started
133
134## 1.0.59
135
136- SDK: Added tool confirmation support with canUseTool callback
137- SDK: Allow specifying env for spawned process
138- Hooks: Exposed PermissionDecision to hooks (including "ask")
139- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
140- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
141
142## 1.0.58
143
144- Added support for reading PDFs
145- MCP: Improved server health status display in 'claude mcp list'
146- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
147
148## 1.0.57
149
150- Added support for specifying a model in slash commands
151- Improved permission messages to help Claude understand allowed tools
152- Fix: Remove trailing newlines from bash output in terminal wrapping
153
154## 1.0.56
155
156- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
157- Fixes for WSL IDE detection
158- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
159
160## 1.0.55
161
162- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
163- Windows: fixed Ctrl+Z crash
164- SDK: Added ability to capture error logging
165- Add --system-prompt-file option to override system prompt in print mode
166
167## 1.0.54
168
169- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
170- Custom slash commands: Added argument-hint to frontmatter
171- Windows: OAuth uses port 45454 and properly constructs browser URL
172- Windows: mode switching now uses alt + m, and plan mode renders properly
173- Shell: Switch to in-memory shell snapshot to fix file-related errors
174
175## 1.0.53
176
177- Updated @-mention file truncation from 100 lines to 2000 lines
178- 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).
179
180## 1.0.52
181
182- Added support for MCP server instructions
183
184## 1.0.51
185
186- Added support for native Windows (requires Git for Windows)
187- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
188- Settings: /doctor can now help you identify and fix invalid setting files
189- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
190- Increased auto-compact warning threshold from 60% to 80%
191- Fixed an issue with handling user directories with spaces for shell snapshots
192- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
193- Custom slash commands: Fixed user-level commands in subdirectories
194- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
195
196## 1.0.48
197
198- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
199- Added progress messages to Bash tool based on the last 5 lines of command output
200- Added expanding variables support for MCP server configuration
201- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
202- Improved IDE extension path handling when Claude Code runs in WSL
203- Hooks: Added a PreCompact hook
204- Vim mode: Added c, f/F, t/T
205
206## 1.0.45
207
208- Redesigned Search (Grep) tool with new tool input parameters and features
209- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
210- Fixed config file corruption issue by enforcing atomic writes
211- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
212- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
213- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
214
215## 1.0.44
216
217- New /export command lets you quickly export a conversation for sharing
218- MCP: resource_link tool results are now supported
219- MCP: tool annotations and tool titles now display in /mcp view
220- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
221
222## 1.0.43
223
224- Fixed a bug where the theme selector was saving excessively
225- Hooks: Added EPIPE system error handling
226
227## 1.0.42
228
229- Added tilde (`~`) expansion support to `/add-dir` command
230
231## 1.0.41
232
233- Hooks: Split Stop hook triggering into Stop and SubagentStop
234- Hooks: Enabled optional timeout configuration for each command
235- Hooks: Added "hook_event_name" to hook input
236- Fixed a bug where MCP tools would display twice in tool list
237- New tool parameters JSON for Bash tool in `tool_decision` event
238
239## 1.0.40
240
241- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
242
243## 1.0.39
244
245- New Active Time metric in OpenTelemetry logging
246
247## 1.0.38
248
249- Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://docs.anthropic.com/en/docs/claude-code/hooks
250
251## 1.0.37
252
253- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
254
255## 1.0.36
256
257- Web search now takes today's date into context
258- Fixed a bug where stdio MCP servers were not terminating properly on exit
259
260## 1.0.35
261
262- Added support for MCP OAuth Authorization Server discovery
263
264## 1.0.34
265
266- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
267
268## 1.0.33
269
270- Improved logging functionality with session ID support
271- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
272- Improvements to plan mode
273
274## 1.0.32
275
276- Updated loopback config for litellm
277- Added forceLoginMethod setting to bypass login selection screen
278
279## 1.0.31
280
281- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
282
283## 1.0.30
284
285- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
286- Improved file path autocomplete with filename matching
287- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
288- Enhanced jq regex support for complex filters with pipes and select
289
290## 1.0.29
291
292- Improved CJK character support in cursor navigation and rendering
293
294## 1.0.28
295
296- Slash commands: Fix selector display during history navigation
297- Resizes images before upload to prevent API size limit errors
298- Added XDG_CONFIG_HOME support to configuration directory
299- Performance optimizations for memory usage
300- New attributes (terminal.type, language) in OpenTelemetry logging
301
302## 1.0.27
303
304- Streamable HTTP MCP servers are now supported
305- Remote MCP servers (SSE and HTTP) now support OAuth
306- MCP resources can now be @-mentioned
307- /resume slash command to switch conversations within Claude Code
308
309## 1.0.25
310
311- Slash commands: moved "project" and "user" prefixes to descriptions
312- Slash commands: improved reliability for command discovery
313- Improved support for Ghostty
314- Improved web search reliability
315
316## 1.0.24
317
318- Improved /mcp output
319- Fixed a bug where settings arrays got overwritten instead of merged
320
321## 1.0.23
322
323- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
324- Released Python SDK: pip install claude-code-sdk to get started
325
326## 1.0.22
327
328- SDK: Renamed `total_cost` to `total_cost_usd`
329
330## 1.0.21
331
332- Improved editing of files with tab-based indentation
333- Fix for tool_use without matching tool_result errors
334- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
335
336## 1.0.18
337
338- Added --add-dir CLI argument for specifying additional working directories
339- Added streaming input support without require -p flag
340- Improved startup performance and session storage performance
341- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
342- Added detailed MCP server tools display (/mcp)
343- MCP authentication and permission improvements
344- Added auto-reconnection for MCP SSE connections on disconnect
345- Fixed issue where pasted content was lost when dialogs appeared
346
347## 1.0.17
348
349- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
350- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
351- MCP server list UI improvements
352- Update Claude Code process title to display "claude" instead of "node"
353
354## 1.0.11
355
356- Claude Code can now also be used with a Claude Pro subscription
357- Added /upgrade for smoother switching to Claude Max plans
358- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
359- Improved shell configuration error handling
360- Improved todo list handling during compaction
361
362## 1.0.10
363
364- Added markdown table support
365- Improved streaming performance
366
367## 1.0.8
368
369- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
370- Increased default otel interval from 1s -> 5s
371- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
372- Fixed a regression where search tools unnecessarily asked for permissions
373- Added support for triggering thinking non-English languages
374- Improved compacting UI
375
376## 1.0.7
377
378- Renamed /allowed-tools -> /permissions
379- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
380- Deprecated claude config commands in favor of editing settings.json
381- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
382- Improved error handling for /install-github-app
383- Bugfixes, UI polish, and tool reliability improvements
384
385## 1.0.6
386
387- Improved edit reliability for tab-indented files
388- Respect CLAUDE_CONFIG_DIR everywhere
389- Reduced unnecessary tool permission prompts
390- Added support for symlinks in @file typeahead
391- Bugfixes, UI polish, and tool reliability improvements
392
393## 1.0.4
394
395- Fixed a bug where MCP tool errors weren't being parsed correctly
396
397## 1.0.1
398
399- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
400- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
401- Updated documentation links and OAuth process descriptions
402
403## 1.0.0
404
405- Claude Code is now generally available
406- Introducing Sonnet 4 and Opus 4 models
407
408## 0.2.125
409
410- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
411- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
412
413## 0.2.117
414
415- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
416- Introduced settings.cleanupPeriodDays
417- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
418- Introduced --debug mode
419
420## 0.2.108
421
422- You can now send messages to Claude while it works to steer Claude in real-time
423- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
424- Fixed a bug where thinking was not working in -p mode
425- Fixed a regression in /cost reporting
426- Deprecated MCP wizard interface in favor of other MCP commands
427- Lots of other bugfixes and improvements
428
429## 0.2.107
430
431- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
432
433## 0.2.106
434
435- MCP SSE server configs can now specify custom headers
436- Fixed a bug where MCP permission prompt didn't always show correctly
437
438## 0.2.105
439
440- Claude can now search the web
441- Moved system & account status to /status
442- Added word movement keybindings for Vim
443- Improved latency for startup, todo tool, and file edits
444
445## 0.2.102
446
447- Improved thinking triggering reliability
448- Improved @mention reliability for images and folders
449- You can now paste multiple large chunks into one prompt
450
451## 0.2.100
452
453- Fixed a crash caused by a stack overflow error
454- Made db storage optional; missing db support disables --continue and --resume
455
456## 0.2.98
457
458- Fixed an issue where auto-compact was running twice
459
460## 0.2.96
461
462- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
463
464## 0.2.93
465
466- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
467- Claude now has access to a Todo list that helps it stay on track and be more organized
468
469## 0.2.82
470
471- Added support for --disallowedTools
472- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
473
474## 0.2.75
475
476- Hit Enter to queue up additional messages while Claude is working
477- Drag in or copy/paste image files directly into the prompt
478- @-mention files to directly add them to context
479- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
480- Improved performance for filename auto-complete
481
482## 0.2.74
483
484- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
485- Task tool can now perform writes and run bash commands
486
487## 0.2.72
488
489- Updated spinner to indicate tokens loaded and tool usage
490
491## 0.2.70
492
493- Network commands like curl are now available for Claude to use
494- Claude can now run multiple web queries in parallel
495- Pressing ESC once immediately interrupts Claude in Auto-accept mode
496
497## 0.2.69
498
499- Fixed UI glitches with improved Select component behavior
500- Enhanced terminal output display with better text truncation logic
501
502## 0.2.67
503
504- Shared project permission rules can be saved in .claude/settings.json
505
506## 0.2.66
507
508- Print mode (-p) now supports streaming output via --output-format=stream-json
509- Fixed issue where pasting could trigger memory or bash mode unexpectedly
510
511## 0.2.63
512
513- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
514
515## 0.2.61
516
517- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
518- Enhanced image detection for more reliable clipboard paste functionality
519- Fixed an issue where ESC key could crash the conversation history selector
520
521## 0.2.59
522
523- Copy+paste images directly into your prompt
524- Improved progress indicators for bash and fetch tools
525- Bugfixes for non-interactive mode (-p)
526
527## 0.2.54
528
529- Quickly add to Memory by starting your message with '#'
530- Press ctrl+r to see full output for long tool results
531- Added support for MCP SSE transport
532
533## 0.2.53
534
535- New web fetch tool lets Claude view URLs that you paste in
536- Fixed a bug with JPEG detection
537
538## 0.2.50
539
540- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
541
542## 0.2.49
543
544- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
545
546## 0.2.47
547
548- Press Tab to auto-complete file and folder names
549- Press Shift + Tab to toggle auto-accept for file edits
550- Automatic conversation compaction for infinite conversation length (toggle with /config)
551
552## 0.2.44
553
554- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
555
556## 0.2.41
557
558- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
559- MCP server startup no longer blocks the app from starting up
560
561## 0.2.37
562
563- New /release-notes command lets you view release notes at any time
564- `claude config add/remove` commands now accept multiple values separated by commas or spaces
565
566## 0.2.36
567
568- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
569- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
570
571## 0.2.34
572
573- Vim bindings for text input - enable with /vim or /config
574
575## 0.2.32
576
577- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
578- Fix for some PersistentShell issues
579
580## 0.2.31
581
582- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
583- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
584
585## 0.2.30
586
587- Added ANSI color theme for better terminal compatibility
588- Fixed issue where slash command arguments weren't being sent properly
589- (Mac-only) API keys are now stored in macOS Keychain
590
591## 0.2.26
592
593- New /approved-tools command for managing tool permissions
594- Word-level diff display for improved code readability
595- Fuzzy matching for slash commands
596
597## 0.2.21
598
599- Fuzzy matching for /commands