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 f54569e · on Aug 14, 2025 History
560 lines · 19.6 KBPreviewCodeRaw
1# Changelog
2
3## 1.0.81
4
5- Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.anthropic.com/en/docs/claude-code/output-styles
6- Agents: Fix custom agent loading when agent files are unparsable
7
8## 1.0.80
9
10- UI improvements: Fix text contrast for custom subagent colors and spinner rendering issues
11
12## 1.0.77
13
14- Bash tool: Fix heredoc and multiline string escaping, improve stderr redirection handling
15- SDK: Add session support and permission denial tracking
16- Fix token limit errors in conversation summarization
17- Opus Plan Mode: New setting in `/model` to run Opus only in plan mode, Sonnet otherwise
18
19## 1.0.73
20
21- MCP: Support multiple config files with `--mcp-config file1.json file2.json`
22- MCP: Press Esc to cancel OAuth authentication flows
23- Bash: Improved command validation and reduced false security warnings
24- UI: Enhanced spinner animations and status line visual hierarchy
25- Linux: Added support for Alpine and musl-based distributions (requires separate ripgrep installation)
26
27## 1.0.72
28
29- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
30
31## 1.0.71
32
33- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
34- Customizable status line: add your terminal prompt to Claude Code with /statusline
35
36## 1.0.70
37
38- Performance: Optimized message rendering for better performance with large contexts
39- Windows: Fixed native file search, ripgrep, and subagent functionality
40- Added support for @-mentions in slash command arguments
41
42## 1.0.69
43
44- Upgraded Opus to version 4.1
45
46## 1.0.68
47
48- Fix incorrect model names being used for certain commands like `/pr-comments`
49- 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.
50- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
51- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
52- SDK: Added canUseTool callback support for tool confirmation
53- Added `disableAllHooks` setting
54- Improved file suggestions performance in large repos
55
56## 1.0.65
57
58- IDE: Fixed connection stability issues and error handling for diagnostics
59- Windows: Fixed shell environment setup for users without .bashrc files
60
61## 1.0.64
62
63- Agents: Added model customization support - you can now specify which model an agent should use
64- Agents: Fixed unintended access to the recursive agent tool
65- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
66- SDK: Fixed user input tracking across multi-turn conversations
67- Added hidden files to file search and @-mention suggestions
68
69## 1.0.63
70
71- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
72
73## 1.0.62
74
75- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
76- Hooks: Added SessionStart hook for new session initialization
77- /add-dir command now supports typeahead for directory paths
78- Improved network connectivity check reliability
79
80## 1.0.61
81
82- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
83- Settings: Added `--settings` flag to load settings from a JSON file
84- Settings: Fixed resolution of settings files paths that are symlinks
85- OTEL: Fixed reporting of wrong organization after authentication changes
86- Slash commands: Fixed permissions checking for allowed-tools with Bash
87- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
88- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
89- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
90
91## 1.0.60
92
93- You can now create custom subagents for specialized tasks! Run /agents to get started
94
95## 1.0.59
96
97- SDK: Added tool confirmation support with canUseTool callback
98- SDK: Allow specifying env for spawned process
99- Hooks: Exposed PermissionDecision to hooks (including "ask")
100- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
101- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
102
103## 1.0.58
104
105- Added support for reading PDFs
106- MCP: Improved server health status display in 'claude mcp list'
107- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
108
109## 1.0.57
110
111- Added support for specifying a model in slash commands
112- Improved permission messages to help Claude understand allowed tools
113- Fix: Remove trailing newlines from bash output in terminal wrapping
114
115## 1.0.56
116
117- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
118- Fixes for WSL IDE detection
119- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
120
121## 1.0.55
122
123- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
124- Windows: fixed Ctrl+Z crash
125- SDK: Added ability to capture error logging
126- Add --system-prompt-file option to override system prompt in print mode
127
128## 1.0.54
129
130- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
131- Custom slash commands: Added argument-hint to frontmatter
132- Windows: OAuth uses port 45454 and properly constructs browser URL
133- Windows: mode switching now uses alt + m, and plan mode renders properly
134- Shell: Switch to in-memory shell snapshot to fix file-related errors
135
136## 1.0.53
137
138- Updated @-mention file truncation from 100 lines to 2000 lines
139- 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).
140
141## 1.0.52
142
143- Added support for MCP server instructions
144
145## 1.0.51
146
147- Added support for native Windows (requires Git for Windows)
148- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
149- Settings: /doctor can now help you identify and fix invalid setting files
150- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
151- Increased auto-compact warning threshold from 60% to 80%
152- Fixed an issue with handling user directories with spaces for shell snapshots
153- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
154- Custom slash commands: Fixed user-level commands in subdirectories
155- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
156
157## 1.0.48
158
159- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
160- Added progress messages to Bash tool based on the last 5 lines of command output
161- Added expanding variables support for MCP server configuration
162- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
163- Improved IDE extension path handling when Claude Code runs in WSL
164- Hooks: Added a PreCompact hook
165- Vim mode: Added c, f/F, t/T
166
167## 1.0.45
168
169- Redesigned Search (Grep) tool with new tool input parameters and features
170- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
171- Fixed config file corruption issue by enforcing atomic writes
172- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
173- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
174- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
175
176## 1.0.44
177
178- New /export command lets you quickly export a conversation for sharing
179- MCP: resource_link tool results are now supported
180- MCP: tool annotations and tool titles now display in /mcp view
181- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
182
183## 1.0.43
184
185- Fixed a bug where the theme selector was saving excessively
186- Hooks: Added EPIPE system error handling
187
188## 1.0.42
189
190- Added tilde (`~`) expansion support to `/add-dir` command
191
192## 1.0.41
193
194- Hooks: Split Stop hook triggering into Stop and SubagentStop
195- Hooks: Enabled optional timeout configuration for each command
196- Hooks: Added "hook_event_name" to hook input
197- Fixed a bug where MCP tools would display twice in tool list
198- New tool parameters JSON for Bash tool in `tool_decision` event
199
200## 1.0.40
201
202- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
203
204## 1.0.39
205
206- New Active Time metric in OpenTelemetry logging
207
208## 1.0.38
209
210- 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
211
212## 1.0.37
213
214- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
215
216## 1.0.36
217
218- Web search now takes today's date into context
219- Fixed a bug where stdio MCP servers were not terminating properly on exit
220
221## 1.0.35
222
223- Added support for MCP OAuth Authorization Server discovery
224
225## 1.0.34
226
227- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
228
229## 1.0.33
230
231- Improved logging functionality with session ID support
232- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
233- Improvements to plan mode
234
235## 1.0.32
236
237- Updated loopback config for litellm
238- Added forceLoginMethod setting to bypass login selection screen
239
240## 1.0.31
241
242- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
243
244## 1.0.30
245
246- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
247- Improved file path autocomplete with filename matching
248- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
249- Enhanced jq regex support for complex filters with pipes and select
250
251## 1.0.29
252
253- Improved CJK character support in cursor navigation and rendering
254
255## 1.0.28
256
257- Slash commands: Fix selector display during history navigation
258- Resizes images before upload to prevent API size limit errors
259- Added XDG_CONFIG_HOME support to configuration directory
260- Performance optimizations for memory usage
261- New attributes (terminal.type, language) in OpenTelemetry logging
262
263## 1.0.27
264
265- Streamable HTTP MCP servers are now supported
266- Remote MCP servers (SSE and HTTP) now support OAuth
267- MCP resources can now be @-mentioned
268- /resume slash command to switch conversations within Claude Code
269
270## 1.0.25
271
272- Slash commands: moved "project" and "user" prefixes to descriptions
273- Slash commands: improved reliability for command discovery
274- Improved support for Ghostty
275- Improved web search reliability
276
277## 1.0.24
278
279- Improved /mcp output
280- Fixed a bug where settings arrays got overwritten instead of merged
281
282## 1.0.23
283
284- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
285- Released Python SDK: pip install claude-code-sdk to get started
286
287## 1.0.22
288
289- SDK: Renamed `total_cost` to `total_cost_usd`
290
291## 1.0.21
292
293- Improved editing of files with tab-based indentation
294- Fix for tool_use without matching tool_result errors
295- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
296
297## 1.0.18
298
299- Added --add-dir CLI argument for specifying additional working directories
300- Added streaming input support without require -p flag
301- Improved startup performance and session storage performance
302- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
303- Added detailed MCP server tools display (/mcp)
304- MCP authentication and permission improvements
305- Added auto-reconnection for MCP SSE connections on disconnect
306- Fixed issue where pasted content was lost when dialogs appeared
307
308## 1.0.17
309
310- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
311- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
312- MCP server list UI improvements
313- Update Claude Code process title to display "claude" instead of "node"
314
315## 1.0.11
316
317- Claude Code can now also be used with a Claude Pro subscription
318- Added /upgrade for smoother switching to Claude Max plans
319- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
320- Improved shell configuration error handling
321- Improved todo list handling during compaction
322
323## 1.0.10
324
325- Added markdown table support
326- Improved streaming performance
327
328## 1.0.8
329
330- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
331- Increased default otel interval from 1s -> 5s
332- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
333- Fixed a regression where search tools unnecessarily asked for permissions
334- Added support for triggering thinking non-English languages
335- Improved compacting UI
336
337## 1.0.7
338
339- Renamed /allowed-tools -> /permissions
340- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
341- Deprecated claude config commands in favor of editing settings.json
342- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
343- Improved error handling for /install-github-app
344- Bugfixes, UI polish, and tool reliability improvements
345
346## 1.0.6
347
348- Improved edit reliability for tab-indented files
349- Respect CLAUDE_CONFIG_DIR everywhere
350- Reduced unnecessary tool permission prompts
351- Added support for symlinks in @file typeahead
352- Bugfixes, UI polish, and tool reliability improvements
353
354## 1.0.4
355
356- Fixed a bug where MCP tool errors weren't being parsed correctly
357
358## 1.0.1
359
360- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
361- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
362- Updated documentation links and OAuth process descriptions
363
364## 1.0.0
365
366- Claude Code is now generally available
367- Introducing Sonnet 4 and Opus 4 models
368
369## 0.2.125
370
371- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
372- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
373
374## 0.2.117
375
376- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
377- Introduced settings.cleanupPeriodDays
378- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
379- Introduced --debug mode
380
381## 0.2.108
382
383- You can now send messages to Claude while it works to steer Claude in real-time
384- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
385- Fixed a bug where thinking was not working in -p mode
386- Fixed a regression in /cost reporting
387- Deprecated MCP wizard interface in favor of other MCP commands
388- Lots of other bugfixes and improvements
389
390## 0.2.107
391
392- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
393
394## 0.2.106
395
396- MCP SSE server configs can now specify custom headers
397- Fixed a bug where MCP permission prompt didn't always show correctly
398
399## 0.2.105
400
401- Claude can now search the web
402- Moved system & account status to /status
403- Added word movement keybindings for Vim
404- Improved latency for startup, todo tool, and file edits
405
406## 0.2.102
407
408- Improved thinking triggering reliability
409- Improved @mention reliability for images and folders
410- You can now paste multiple large chunks into one prompt
411
412## 0.2.100
413
414- Fixed a crash caused by a stack overflow error
415- Made db storage optional; missing db support disables --continue and --resume
416
417## 0.2.98
418
419- Fixed an issue where auto-compact was running twice
420
421## 0.2.96
422
423- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
424
425## 0.2.93
426
427- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
428- Claude now has access to a Todo list that helps it stay on track and be more organized
429
430## 0.2.82
431
432- Added support for --disallowedTools
433- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
434
435## 0.2.75
436
437- Hit Enter to queue up additional messages while Claude is working
438- Drag in or copy/paste image files directly into the prompt
439- @-mention files to directly add them to context
440- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
441- Improved performance for filename auto-complete
442
443## 0.2.74
444
445- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
446- Task tool can now perform writes and run bash commands
447
448## 0.2.72
449
450- Updated spinner to indicate tokens loaded and tool usage
451
452## 0.2.70
453
454- Network commands like curl are now available for Claude to use
455- Claude can now run multiple web queries in parallel
456- Pressing ESC once immediately interrupts Claude in Auto-accept mode
457
458## 0.2.69
459
460- Fixed UI glitches with improved Select component behavior
461- Enhanced terminal output display with better text truncation logic
462
463## 0.2.67
464
465- Shared project permission rules can be saved in .claude/settings.json
466
467## 0.2.66
468
469- Print mode (-p) now supports streaming output via --output-format=stream-json
470- Fixed issue where pasting could trigger memory or bash mode unexpectedly
471
472## 0.2.63
473
474- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
475
476## 0.2.61
477
478- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
479- Enhanced image detection for more reliable clipboard paste functionality
480- Fixed an issue where ESC key could crash the conversation history selector
481
482## 0.2.59
483
484- Copy+paste images directly into your prompt
485- Improved progress indicators for bash and fetch tools
486- Bugfixes for non-interactive mode (-p)
487
488## 0.2.54
489
490- Quickly add to Memory by starting your message with '#'
491- Press ctrl+r to see full output for long tool results
492- Added support for MCP SSE transport
493
494## 0.2.53
495
496- New web fetch tool lets Claude view URLs that you paste in
497- Fixed a bug with JPEG detection
498
499## 0.2.50
500
501- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
502
503## 0.2.49
504
505- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
506
507## 0.2.47
508
509- Press Tab to auto-complete file and folder names
510- Press Shift + Tab to toggle auto-accept for file edits
511- Automatic conversation compaction for infinite conversation length (toggle with /config)
512
513## 0.2.44
514
515- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
516
517## 0.2.41
518
519- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
520- MCP server startup no longer blocks the app from starting up
521
522## 0.2.37
523
524- New /release-notes command lets you view release notes at any time
525- `claude config add/remove` commands now accept multiple values separated by commas or spaces
526
527## 0.2.36
528
529- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
530- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
531
532## 0.2.34
533
534- Vim bindings for text input - enable with /vim or /config
535
536## 0.2.32
537
538- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
539- Fix for some PersistentShell issues
540
541## 0.2.31
542
543- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
544- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
545
546## 0.2.30
547
548- Added ANSI color theme for better terminal compatibility
549- Fixed issue where slash command arguments weren't being sent properly
550- (Mac-only) API keys are now stored in macOS Keychain
551
552## 0.2.26
553
554- New /approved-tools command for managing tool permissions
555- Word-level diff display for improved code readability
556- Fuzzy matching for slash commands
557
558## 0.2.21
559
560- Fuzzy matching for /commands