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 81f65be · on Aug 29, 2025 History
619 lines · 21.6 KBPreviewCodeRaw
1# Changelog
2
3## 1.0.97
4
5- Settings: /doctor now validates permission rule syntax and suggests corrections
6
7## 1.0.94
8
9- Vertex: add support for global endpoints for supported models
10- /memory command now allows direct editing of all imported memory files
11- SDK: Add custom tools as callbacks
12- Added /todos command to list current todo items
13
14## 1.0.93
15
16- Windows: Add alt + v shortcut for pasting images from clipboard
17- Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
18
19## 1.0.90
20
21- Settings file changes take effect immediately - no restart required
22
23## 1.0.88
24
25- Fixed issue causing "OAuth authentication is currently not supported"
26- Status line input now includes `exceeds_200k_tokens`
27- Fixed incorrect usage tracking in /cost.
28- Introduced `ANTHROPIC_DEFAULT_SONNET_MODEL` and `ANTHROPIC_DEFAULT_OPUS_MODEL` for controlling model aliases opusplan, opus, and sonnet.
29- Bedrock: Updated default Sonnet model to Sonnet 4
30
31## 1.0.86
32
33- Added /context to help users self-serve debug context issues
34- SDK: Added UUID support for all SDK messages
35- SDK: Added `--replay-user-messages` to replay user messages back to stdout
36
37## 1.0.85
38
39- Status line input now includes session cost info
40- Hooks: Introduced SessionEnd hook
41
42## 1.0.84
43
44- Fix tool_use/tool_result id mismatch error when network is unstable
45- Fix Claude sometimes ignoring real-time steering when wrapping up a task
46- @-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
47- Use built-in ripgrep by default; to opt out of this behavior, set USE_BUILTIN_RIPGREP=0
48
49## 1.0.83
50
51- @-mention: Support files with spaces in path
52- New shimmering spinner
53
54## 1.0.82
55
56- SDK: Add request cancellation support
57- SDK: New additionalDirectories option to search custom paths, improved slash command processing
58- Settings: Validation prevents invalid fields in .claude/settings.json files
59- MCP: Improve tool name consistency
60- Bash: Fix crash when Claude tries to automatically read large files
61
62## 1.0.81
63
64- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.anthropic.com/en/docs/claude-code/output-styles
65- Agents: Fix custom agent loading when agent files are unparsable
66
67## 1.0.80
68
69- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
70
71## 1.0.77
72
73- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
74- SDK: Add session support and permission denial tracking
75- Fix token limit errors in conversation summarization
76- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
77
78## 1.0.73
79
80- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
81- MCP: Press Esc to cancel OAuth authentication flows
82- Bash: Improved command validation and reduced false security warnings
83- UI: Enhanced spinner animations and status line visual hierarchy
84- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
85
86## 1.0.72
87
88- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
89
90## 1.0.71
91
92- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
93- Customizable status line: add your terminal prompt to Claude Code with /statusline
94
95## 1.0.70
96
97- Performance: Optimized message rendering for better performance with large contexts
98- Windows: Fixed native file search, ripgrep, and subagent functionality
99- Added support for @-mentions in slash command arguments
100
101## 1.0.69
102
103- Upgraded Opus to version 4.1
104
105## 1.0.68
106
107- Fix incorrect model names being used for certain commands like `/pr-comments`
108- 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.
109- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
110- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
111- SDK: Added canUseTool callback support for tool confirmation
112- Added `disableAllHooks` setting
113- Improved file suggestions performance in large repos
114
115## 1.0.65
116
117- IDE: Fixed connection stability issues and error handling for diagnostics
118- Windows: Fixed shell environment setup for users without .bashrc files
119
120## 1.0.64
121
122- Agents: Added model customization support - you can now specify which model an agent should use
123- Agents: Fixed unintended access to the recursive agent tool
124- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
125- SDK: Fixed user input tracking across multi-turn conversations
126- Added hidden files to file search and @-mention suggestions
127
128## 1.0.63
129
130- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
131
132## 1.0.62
133
134- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
135- Hooks: Added SessionStart hook for new session initialization
136- /add-dir command now supports typeahead for directory paths
137- Improved network connectivity check reliability
138
139## 1.0.61
140
141- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
142- Settings: Added `--settings` flag to load settings from a JSON file
143- Settings: Fixed resolution of settings files paths that are symlinks
144- OTEL: Fixed reporting of wrong organization after authentication changes
145- Slash commands: Fixed permissions checking for allowed-tools with Bash
146- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
147- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
148- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
149
150## 1.0.60
151
152- You can now create custom subagents for specialized tasks! Run /agents to get started
153
154## 1.0.59
155
156- SDK: Added tool confirmation support with canUseTool callback
157- SDK: Allow specifying env for spawned process
158- Hooks: Exposed PermissionDecision to hooks (including "ask")
159- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
160- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
161
162## 1.0.58
163
164- Added support for reading PDFs
165- MCP: Improved server health status display in 'claude mcp list'
166- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
167
168## 1.0.57
169
170- Added support for specifying a model in slash commands
171- Improved permission messages to help Claude understand allowed tools
172- Fix: Remove trailing newlines from bash output in terminal wrapping
173
174## 1.0.56
175
176- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
177- Fixes for WSL IDE detection
178- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
179
180## 1.0.55
181
182- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
183- Windows: fixed Ctrl+Z crash
184- SDK: Added ability to capture error logging
185- Add --system-prompt-file option to override system prompt in print mode
186
187## 1.0.54
188
189- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
190- Custom slash commands: Added argument-hint to frontmatter
191- Windows: OAuth uses port 45454 and properly constructs browser URL
192- Windows: mode switching now uses alt + m, and plan mode renders properly
193- Shell: Switch to in-memory shell snapshot to fix file-related errors
194
195## 1.0.53
196
197- Updated @-mention file truncation from 100 lines to 2000 lines
198- 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).
199
200## 1.0.52
201
202- Added support for MCP server instructions
203
204## 1.0.51
205
206- Added support for native Windows (requires Git for Windows)
207- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
208- Settings: /doctor can now help you identify and fix invalid setting files
209- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
210- Increased auto-compact warning threshold from 60% to 80%
211- Fixed an issue with handling user directories with spaces for shell snapshots
212- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
213- Custom slash commands: Fixed user-level commands in subdirectories
214- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
215
216## 1.0.48
217
218- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
219- Added progress messages to Bash tool based on the last 5 lines of command output
220- Added expanding variables support for MCP server configuration
221- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
222- Improved IDE extension path handling when Claude Code runs in WSL
223- Hooks: Added a PreCompact hook
224- Vim mode: Added c, f/F, t/T
225
226## 1.0.45
227
228- Redesigned Search (Grep) tool with new tool input parameters and features
229- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
230- Fixed config file corruption issue by enforcing atomic writes
231- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
232- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
233- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
234
235## 1.0.44
236
237- New /export command lets you quickly export a conversation for sharing
238- MCP: resource_link tool results are now supported
239- MCP: tool annotations and tool titles now display in /mcp view
240- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
241
242## 1.0.43
243
244- Fixed a bug where the theme selector was saving excessively
245- Hooks: Added EPIPE system error handling
246
247## 1.0.42
248
249- Added tilde (`~`) expansion support to `/add-dir` command
250
251## 1.0.41
252
253- Hooks: Split Stop hook triggering into Stop and SubagentStop
254- Hooks: Enabled optional timeout configuration for each command
255- Hooks: Added "hook_event_name" to hook input
256- Fixed a bug where MCP tools would display twice in tool list
257- New tool parameters JSON for Bash tool in `tool_decision` event
258
259## 1.0.40
260
261- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
262
263## 1.0.39
264
265- New Active Time metric in OpenTelemetry logging
266
267## 1.0.38
268
269- 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
270
271## 1.0.37
272
273- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
274
275## 1.0.36
276
277- Web search now takes today's date into context
278- Fixed a bug where stdio MCP servers were not terminating properly on exit
279
280## 1.0.35
281
282- Added support for MCP OAuth Authorization Server discovery
283
284## 1.0.34
285
286- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
287
288## 1.0.33
289
290- Improved logging functionality with session ID support
291- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
292- Improvements to plan mode
293
294## 1.0.32
295
296- Updated loopback config for litellm
297- Added forceLoginMethod setting to bypass login selection screen
298
299## 1.0.31
300
301- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
302
303## 1.0.30
304
305- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
306- Improved file path autocomplete with filename matching
307- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
308- Enhanced jq regex support for complex filters with pipes and select
309
310## 1.0.29
311
312- Improved CJK character support in cursor navigation and rendering
313
314## 1.0.28
315
316- Slash commands: Fix selector display during history navigation
317- Resizes images before upload to prevent API size limit errors
318- Added XDG_CONFIG_HOME support to configuration directory
319- Performance optimizations for memory usage
320- New attributes (terminal.type, language) in OpenTelemetry logging
321
322## 1.0.27
323
324- Streamable HTTP MCP servers are now supported
325- Remote MCP servers (SSE and HTTP) now support OAuth
326- MCP resources can now be @-mentioned
327- /resume slash command to switch conversations within Claude Code
328
329## 1.0.25
330
331- Slash commands: moved "project" and "user" prefixes to descriptions
332- Slash commands: improved reliability for command discovery
333- Improved support for Ghostty
334- Improved web search reliability
335
336## 1.0.24
337
338- Improved /mcp output
339- Fixed a bug where settings arrays got overwritten instead of merged
340
341## 1.0.23
342
343- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
344- Released Python SDK: pip install claude-code-sdk to get started
345
346## 1.0.22
347
348- SDK: Renamed `total_cost` to `total_cost_usd`
349
350## 1.0.21
351
352- Improved editing of files with tab-based indentation
353- Fix for tool_use without matching tool_result errors
354- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
355
356## 1.0.18
357
358- Added --add-dir CLI argument for specifying additional working directories
359- Added streaming input support without require -p flag
360- Improved startup performance and session storage performance
361- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
362- Added detailed MCP server tools display (/mcp)
363- MCP authentication and permission improvements
364- Added auto-reconnection for MCP SSE connections on disconnect
365- Fixed issue where pasted content was lost when dialogs appeared
366
367## 1.0.17
368
369- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
370- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
371- MCP server list UI improvements
372- Update Claude Code process title to display "claude" instead of "node"
373
374## 1.0.11
375
376- Claude Code can now also be used with a Claude Pro subscription
377- Added /upgrade for smoother switching to Claude Max plans
378- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
379- Improved shell configuration error handling
380- Improved todo list handling during compaction
381
382## 1.0.10
383
384- Added markdown table support
385- Improved streaming performance
386
387## 1.0.8
388
389- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
390- Increased default otel interval from 1s -> 5s
391- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
392- Fixed a regression where search tools unnecessarily asked for permissions
393- Added support for triggering thinking non-English languages
394- Improved compacting UI
395
396## 1.0.7
397
398- Renamed /allowed-tools -> /permissions
399- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
400- Deprecated claude config commands in favor of editing settings.json
401- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
402- Improved error handling for /install-github-app
403- Bugfixes, UI polish, and tool reliability improvements
404
405## 1.0.6
406
407- Improved edit reliability for tab-indented files
408- Respect CLAUDE_CONFIG_DIR everywhere
409- Reduced unnecessary tool permission prompts
410- Added support for symlinks in @file typeahead
411- Bugfixes, UI polish, and tool reliability improvements
412
413## 1.0.4
414
415- Fixed a bug where MCP tool errors weren't being parsed correctly
416
417## 1.0.1
418
419- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
420- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
421- Updated documentation links and OAuth process descriptions
422
423## 1.0.0
424
425- Claude Code is now generally available
426- Introducing Sonnet 4 and Opus 4 models
427
428## 0.2.125
429
430- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
431- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
432
433## 0.2.117
434
435- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
436- Introduced settings.cleanupPeriodDays
437- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
438- Introduced --debug mode
439
440## 0.2.108
441
442- You can now send messages to Claude while it works to steer Claude in real-time
443- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
444- Fixed a bug where thinking was not working in -p mode
445- Fixed a regression in /cost reporting
446- Deprecated MCP wizard interface in favor of other MCP commands
447- Lots of other bugfixes and improvements
448
449## 0.2.107
450
451- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
452
453## 0.2.106
454
455- MCP SSE server configs can now specify custom headers
456- Fixed a bug where MCP permission prompt didn't always show correctly
457
458## 0.2.105
459
460- Claude can now search the web
461- Moved system & account status to /status
462- Added word movement keybindings for Vim
463- Improved latency for startup, todo tool, and file edits
464
465## 0.2.102
466
467- Improved thinking triggering reliability
468- Improved @mention reliability for images and folders
469- You can now paste multiple large chunks into one prompt
470
471## 0.2.100
472
473- Fixed a crash caused by a stack overflow error
474- Made db storage optional; missing db support disables --continue and --resume
475
476## 0.2.98
477
478- Fixed an issue where auto-compact was running twice
479
480## 0.2.96
481
482- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
483
484## 0.2.93
485
486- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
487- Claude now has access to a Todo list that helps it stay on track and be more organized
488
489## 0.2.82
490
491- Added support for --disallowedTools
492- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
493
494## 0.2.75
495
496- Hit Enter to queue up additional messages while Claude is working
497- Drag in or copy/paste image files directly into the prompt
498- @-mention files to directly add them to context
499- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
500- Improved performance for filename auto-complete
501
502## 0.2.74
503
504- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
505- Task tool can now perform writes and run bash commands
506
507## 0.2.72
508
509- Updated spinner to indicate tokens loaded and tool usage
510
511## 0.2.70
512
513- Network commands like curl are now available for Claude to use
514- Claude can now run multiple web queries in parallel
515- Pressing ESC once immediately interrupts Claude in Auto-accept mode
516
517## 0.2.69
518
519- Fixed UI glitches with improved Select component behavior
520- Enhanced terminal output display with better text truncation logic
521
522## 0.2.67
523
524- Shared project permission rules can be saved in .claude/settings.json
525
526## 0.2.66
527
528- Print mode (-p) now supports streaming output via --output-format=stream-json
529- Fixed issue where pasting could trigger memory or bash mode unexpectedly
530
531## 0.2.63
532
533- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
534
535## 0.2.61
536
537- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
538- Enhanced image detection for more reliable clipboard paste functionality
539- Fixed an issue where ESC key could crash the conversation history selector
540
541## 0.2.59
542
543- Copy+paste images directly into your prompt
544- Improved progress indicators for bash and fetch tools
545- Bugfixes for non-interactive mode (-p)
546
547## 0.2.54
548
549- Quickly add to Memory by starting your message with '#'
550- Press ctrl+r to see full output for long tool results
551- Added support for MCP SSE transport
552
553## 0.2.53
554
555- New web fetch tool lets Claude view URLs that you paste in
556- Fixed a bug with JPEG detection
557
558## 0.2.50
559
560- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
561
562## 0.2.49
563
564- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
565
566## 0.2.47
567
568- Press Tab to auto-complete file and folder names
569- Press Shift + Tab to toggle auto-accept for file edits
570- Automatic conversation compaction for infinite conversation length (toggle with /config)
571
572## 0.2.44
573
574- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
575
576## 0.2.41
577
578- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
579- MCP server startup no longer blocks the app from starting up
580
581## 0.2.37
582
583- New /release-notes command lets you view release notes at any time
584- `claude config add/remove` commands now accept multiple values separated by commas or spaces
585
586## 0.2.36
587
588- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
589- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
590
591## 0.2.34
592
593- Vim bindings for text input - enable with /vim or /config
594
595## 0.2.32
596
597- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
598- Fix for some PersistentShell issues
599
600## 0.2.31
601
602- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
603- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
604
605## 0.2.30
606
607- Added ANSI color theme for better terminal compatibility
608- Fixed issue where slash command arguments weren't being sent properly
609- (Mac-only) API keys are now stored in macOS Keychain
610
611## 0.2.26
612
613- New /approved-tools command for managing tool permissions
614- Word-level diff display for improved code readability
615- Fuzzy matching for slash commands
616
617## 0.2.21
618
619- Fuzzy matching for /commands