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 dd53f86 · on Aug 11, 2025 History
536 lines · 18.6 KBPreviewCodeRaw
1# Changelog
2
3## 1.0.72
4
5- Ask permissions: have Claude Code always ask for confirmation to use specific tools with /permissions
6
7## 1.0.71
8
9- Background commands: (Ctrl-b) to run any Bash command in the background so Claude can keep working (great for dev servers, tailing logs, etc.)
10- Customizable status line: add your terminal prompt to Claude Code with /statusline
11
12## 1.0.70
13
14- Performance: Optimized message rendering for better performance with large contexts
15- Windows: Fixed native file search, ripgrep, and subagent functionality
16- Added support for @-mentions in slash command arguments
17
18## 1.0.69
19
20- Upgraded Opus to version 4.1
21
22## 1.0.68
23
24- Fix incorrect model names being used for certain commands like `/pr-comments`
25- 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.
26- Windows: improve sub-process spawning to eliminate "No such file or directory" when running commands like pnpm
27- Enhanced /doctor command with CLAUDE.md and MCP tool context for self-serve debugging
28- SDK: Added canUseTool callback support for tool confirmation
29- Added `disableAllHooks` setting
30- Improved file suggestions performance in large repos
31
32## 1.0.65
33
34- IDE: Fixed connection stability issues and error handling for diagnostics
35- Windows: Fixed shell environment setup for users without .bashrc files
36
37## 1.0.64
38
39- Agents: Added model customization support - you can now specify which model an agent should use
40- Agents: Fixed unintended access to the recursive agent tool
41- Hooks: Added systemMessage field to hook JSON output for displaying warnings and context
42- SDK: Fixed user input tracking across multi-turn conversations
43- Added hidden files to file search and @-mention suggestions
44
45## 1.0.63
46
47- Windows: Fixed file search, @agent mentions, and custom slash commands functionality
48
49## 1.0.62
50
51- Added @-mention support with typeahead for custom agents. @<your-custom-agent> to invoke it
52- Hooks: Added SessionStart hook for new session initialization
53- /add-dir command now supports typeahead for directory paths
54- Improved network connectivity check reliability
55
56## 1.0.61
57
58- Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
59- Settings: Added `--settings` flag to load settings from a JSON file
60- Settings: Fixed resolution of settings files paths that are symlinks
61- OTEL: Fixed reporting of wrong organization after authentication changes
62- Slash commands: Fixed permissions checking for allowed-tools with Bash
63- IDE: Added support for pasting images in VSCode MacOS using ⌘+V
64- IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
65- Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
66
67## 1.0.60
68
69- You can now create custom subagents for specialized tasks! Run /agents to get started
70
71## 1.0.59
72
73- SDK: Added tool confirmation support with canUseTool callback
74- SDK: Allow specifying env for spawned process
75- Hooks: Exposed PermissionDecision to hooks (including "ask")
76- Hooks: UserPromptSubmit now supports additionalContext in advanced JSON output
77- Fixed issue where some Max users that specified Opus would still see fallback to Sonnet
78
79## 1.0.58
80
81- Added support for reading PDFs
82- MCP: Improved server health status display in 'claude mcp list'
83- Hooks: Added CLAUDE_PROJECT_DIR env var for hook commands
84
85## 1.0.57
86
87- Added support for specifying a model in slash commands
88- Improved permission messages to help Claude understand allowed tools
89- Fix: Remove trailing newlines from bash output in terminal wrapping
90
91## 1.0.56
92
93- Windows: Enabled shift+tab for mode switching on versions of Node.js that support terminal VT mode
94- Fixes for WSL IDE detection
95- Fix an issue causing awsRefreshHelper changes to .aws directory not to be picked up
96
97## 1.0.55
98
99- Clarified knowledge cutoff for Opus 4 and Sonnet 4 models
100- Windows: fixed Ctrl+Z crash
101- SDK: Added ability to capture error logging
102- Add --system-prompt-file option to override system prompt in print mode
103
104## 1.0.54
105
106- Hooks: Added UserPromptSubmit hook and the current working directory to hook inputs
107- Custom slash commands: Added argument-hint to frontmatter
108- Windows: OAuth uses port 45454 and properly constructs browser URL
109- Windows: mode switching now uses alt + m, and plan mode renders properly
110- Shell: Switch to in-memory shell snapshot to fix file-related errors
111
112## 1.0.53
113
114- Updated @-mention file truncation from 100 lines to 2000 lines
115- 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).
116
117## 1.0.52
118
119- Added support for MCP server instructions
120
121## 1.0.51
122
123- Added support for native Windows (requires Git for Windows)
124- Added support for Bedrock API keys through environment variable AWS_BEARER_TOKEN_BEDROCK
125- Settings: /doctor can now help you identify and fix invalid setting files
126- `--append-system-prompt` can now be used in interactive mode, not just --print/-p.
127- Increased auto-compact warning threshold from 60% to 80%
128- Fixed an issue with handling user directories with spaces for shell snapshots
129- OTEL resource now includes os.type, os.version, host.arch, and wsl.version (if running on Windows Subsystem for Linux)
130- Custom slash commands: Fixed user-level commands in subdirectories
131- Plan mode: Fixed issue where rejected plan from sub-task would get discarded
132
133## 1.0.48
134
135- Fixed a bug in v1.0.45 where the app would sometimes freeze on launch
136- Added progress messages to Bash tool based on the last 5 lines of command output
137- Added expanding variables support for MCP server configuration
138- Moved shell snapshots from /tmp to ~/.claude for more reliable Bash tool calls
139- Improved IDE extension path handling when Claude Code runs in WSL
140- Hooks: Added a PreCompact hook
141- Vim mode: Added c, f/F, t/T
142
143## 1.0.45
144
145- Redesigned Search (Grep) tool with new tool input parameters and features
146- Disabled IDE diffs for notebook files, fixing "Timeout waiting after 1000ms" error
147- Fixed config file corruption issue by enforcing atomic writes
148- Updated prompt input undo to Ctrl+\_ to avoid breaking existing Ctrl+U behavior, matching zsh's undo shortcut
149- Stop Hooks: Fixed transcript path after /clear and fixed triggering when loop ends with tool call
150- Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
151
152## 1.0.44
153
154- New /export command lets you quickly export a conversation for sharing
155- MCP: resource_link tool results are now supported
156- MCP: tool annotations and tool titles now display in /mcp view
157- Changed Ctrl+Z to suspend Claude Code. Resume by running `fg`. Prompt input undo is now Ctrl+U.
158
159## 1.0.43
160
161- Fixed a bug where the theme selector was saving excessively
162- Hooks: Added EPIPE system error handling
163
164## 1.0.42
165
166- Added tilde (`~`) expansion support to `/add-dir` command
167
168## 1.0.41
169
170- Hooks: Split Stop hook triggering into Stop and SubagentStop
171- Hooks: Enabled optional timeout configuration for each command
172- Hooks: Added "hook_event_name" to hook input
173- Fixed a bug where MCP tools would display twice in tool list
174- New tool parameters JSON for Bash tool in `tool_decision` event
175
176## 1.0.40
177
178- Fixed a bug causing API connection errors with UNABLE_TO_GET_ISSUER_CERT_LOCALLY if `NODE_EXTRA_CA_CERTS` was set
179
180## 1.0.39
181
182- New Active Time metric in OpenTelemetry logging
183
184## 1.0.38
185
186- 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
187
188## 1.0.37
189
190- Remove ability to set `Proxy-Authorization` header via ANTHROPIC_AUTH_TOKEN or apiKeyHelper
191
192## 1.0.36
193
194- Web search now takes today's date into context
195- Fixed a bug where stdio MCP servers were not terminating properly on exit
196
197## 1.0.35
198
199- Added support for MCP OAuth Authorization Server discovery
200
201## 1.0.34
202
203- Fixed a memory leak causing a MaxListenersExceededWarning message to appear
204
205## 1.0.33
206
207- Improved logging functionality with session ID support
208- Added prompt input undo functionality (Ctrl+Z and vim 'u' command)
209- Improvements to plan mode
210
211## 1.0.32
212
213- Updated loopback config for litellm
214- Added forceLoginMethod setting to bypass login selection screen
215
216## 1.0.31
217
218- Fixed a bug where ~/.claude.json would get reset when file contained invalid JSON
219
220## 1.0.30
221
222- Custom slash commands: Run bash output, @-mention files, enable thinking with thinking keywords
223- Improved file path autocomplete with filename matching
224- Added timestamps in Ctrl-r mode and fixed Ctrl-c handling
225- Enhanced jq regex support for complex filters with pipes and select
226
227## 1.0.29
228
229- Improved CJK character support in cursor navigation and rendering
230
231## 1.0.28
232
233- Slash commands: Fix selector display during history navigation
234- Resizes images before upload to prevent API size limit errors
235- Added XDG_CONFIG_HOME support to configuration directory
236- Performance optimizations for memory usage
237- New attributes (terminal.type, language) in OpenTelemetry logging
238
239## 1.0.27
240
241- Streamable HTTP MCP servers are now supported
242- Remote MCP servers (SSE and HTTP) now support OAuth
243- MCP resources can now be @-mentioned
244- /resume slash command to switch conversations within Claude Code
245
246## 1.0.25
247
248- Slash commands: moved "project" and "user" prefixes to descriptions
249- Slash commands: improved reliability for command discovery
250- Improved support for Ghostty
251- Improved web search reliability
252
253## 1.0.24
254
255- Improved /mcp output
256- Fixed a bug where settings arrays got overwritten instead of merged
257
258## 1.0.23
259
260- Released TypeScript SDK: import @anthropic-ai/claude-code to get started
261- Released Python SDK: pip install claude-code-sdk to get started
262
263## 1.0.22
264
265- SDK: Renamed `total_cost` to `total_cost_usd`
266
267## 1.0.21
268
269- Improved editing of files with tab-based indentation
270- Fix for tool_use without matching tool_result errors
271- Fixed a bug where stdio MCP server processes would linger after quitting Claude Code
272
273## 1.0.18
274
275- Added --add-dir CLI argument for specifying additional working directories
276- Added streaming input support without require -p flag
277- Improved startup performance and session storage performance
278- Added CLAUDE_BASH_MAINTAIN_PROJECT_WORKING_DIR environment variable to freeze working directory for bash commands
279- Added detailed MCP server tools display (/mcp)
280- MCP authentication and permission improvements
281- Added auto-reconnection for MCP SSE connections on disconnect
282- Fixed issue where pasted content was lost when dialogs appeared
283
284## 1.0.17
285
286- We now emit messages from sub-tasks in -p mode (look for the parent_tool_use_id property)
287- Fixed crashes when the VS Code diff tool is invoked multiple times quickly
288- MCP server list UI improvements
289- Update Claude Code process title to display "claude" instead of "node"
290
291## 1.0.11
292
293- Claude Code can now also be used with a Claude Pro subscription
294- Added /upgrade for smoother switching to Claude Max plans
295- Improved UI for authentication from API keys and Bedrock/Vertex/external auth tokens
296- Improved shell configuration error handling
297- Improved todo list handling during compaction
298
299## 1.0.10
300
301- Added markdown table support
302- Improved streaming performance
303
304## 1.0.8
305
306- Fixed Vertex AI region fallback when using CLOUD_ML_REGION
307- Increased default otel interval from 1s -> 5s
308- Fixed edge cases where MCP_TIMEOUT and MCP_TOOL_TIMEOUT weren't being respected
309- Fixed a regression where search tools unnecessarily asked for permissions
310- Added support for triggering thinking non-English languages
311- Improved compacting UI
312
313## 1.0.7
314
315- Renamed /allowed-tools -> /permissions
316- Migrated allowedTools and ignorePatterns from .claude.json -> settings.json
317- Deprecated claude config commands in favor of editing settings.json
318- Fixed a bug where --dangerously-skip-permissions sometimes didn't work in --print mode
319- Improved error handling for /install-github-app
320- Bugfixes, UI polish, and tool reliability improvements
321
322## 1.0.6
323
324- Improved edit reliability for tab-indented files
325- Respect CLAUDE_CONFIG_DIR everywhere
326- Reduced unnecessary tool permission prompts
327- Added support for symlinks in @file typeahead
328- Bugfixes, UI polish, and tool reliability improvements
329
330## 1.0.4
331
332- Fixed a bug where MCP tool errors weren't being parsed correctly
333
334## 1.0.1
335
336- Added `DISABLE_INTERLEAVED_THINKING` to give users the option to opt out of interleaved thinking.
337- Improved model references to show provider-specific names (Sonnet 3.7 for Bedrock, Sonnet 4 for Console)
338- Updated documentation links and OAuth process descriptions
339
340## 1.0.0
341
342- Claude Code is now generally available
343- Introducing Sonnet 4 and Opus 4 models
344
345## 0.2.125
346
347- Breaking change: Bedrock ARN passed to `ANTHROPIC_MODEL` or `ANTHROPIC_SMALL_FAST_MODEL` should no longer contain an escaped slash (specify `/` instead of `%2F`)
348- Removed `DEBUG=true` in favor of `ANTHROPIC_LOG=debug`, to log all requests
349
350## 0.2.117
351
352- Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
353- Introduced settings.cleanupPeriodDays
354- Introduced CLAUDE_CODE_API_KEY_HELPER_TTL_MS env var
355- Introduced --debug mode
356
357## 0.2.108
358
359- You can now send messages to Claude while it works to steer Claude in real-time
360- Introduced BASH_DEFAULT_TIMEOUT_MS and BASH_MAX_TIMEOUT_MS env vars
361- Fixed a bug where thinking was not working in -p mode
362- Fixed a regression in /cost reporting
363- Deprecated MCP wizard interface in favor of other MCP commands
364- Lots of other bugfixes and improvements
365
366## 0.2.107
367
368- CLAUDE.md files can now import other files. Add @path/to/file.md to ./CLAUDE.md to load additional files on launch
369
370## 0.2.106
371
372- MCP SSE server configs can now specify custom headers
373- Fixed a bug where MCP permission prompt didn't always show correctly
374
375## 0.2.105
376
377- Claude can now search the web
378- Moved system & account status to /status
379- Added word movement keybindings for Vim
380- Improved latency for startup, todo tool, and file edits
381
382## 0.2.102
383
384- Improved thinking triggering reliability
385- Improved @mention reliability for images and folders
386- You can now paste multiple large chunks into one prompt
387
388## 0.2.100
389
390- Fixed a crash caused by a stack overflow error
391- Made db storage optional; missing db support disables --continue and --resume
392
393## 0.2.98
394
395- Fixed an issue where auto-compact was running twice
396
397## 0.2.96
398
399- Claude Code can now also be used with a Claude Max subscription (https://claude.ai/upgrade)
400
401## 0.2.93
402
403- Resume conversations from where you left off from with "claude --continue" and "claude --resume"
404- Claude now has access to a Todo list that helps it stay on track and be more organized
405
406## 0.2.82
407
408- Added support for --disallowedTools
409- Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
410
411## 0.2.75
412
413- Hit Enter to queue up additional messages while Claude is working
414- Drag in or copy/paste image files directly into the prompt
415- @-mention files to directly add them to context
416- Run one-off MCP servers with `claude --mcp-config <path-to-file>`
417- Improved performance for filename auto-complete
418
419## 0.2.74
420
421- Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
422- Task tool can now perform writes and run bash commands
423
424## 0.2.72
425
426- Updated spinner to indicate tokens loaded and tool usage
427
428## 0.2.70
429
430- Network commands like curl are now available for Claude to use
431- Claude can now run multiple web queries in parallel
432- Pressing ESC once immediately interrupts Claude in Auto-accept mode
433
434## 0.2.69
435
436- Fixed UI glitches with improved Select component behavior
437- Enhanced terminal output display with better text truncation logic
438
439## 0.2.67
440
441- Shared project permission rules can be saved in .claude/settings.json
442
443## 0.2.66
444
445- Print mode (-p) now supports streaming output via --output-format=stream-json
446- Fixed issue where pasting could trigger memory or bash mode unexpectedly
447
448## 0.2.63
449
450- Fixed an issue where MCP tools were loaded twice, which caused tool call errors
451
452## 0.2.61
453
454- Navigate menus with vim-style keys (j/k) or bash/emacs shortcuts (Ctrl+n/p) for faster interaction
455- Enhanced image detection for more reliable clipboard paste functionality
456- Fixed an issue where ESC key could crash the conversation history selector
457
458## 0.2.59
459
460- Copy+paste images directly into your prompt
461- Improved progress indicators for bash and fetch tools
462- Bugfixes for non-interactive mode (-p)
463
464## 0.2.54
465
466- Quickly add to Memory by starting your message with '#'
467- Press ctrl+r to see full output for long tool results
468- Added support for MCP SSE transport
469
470## 0.2.53
471
472- New web fetch tool lets Claude view URLs that you paste in
473- Fixed a bug with JPEG detection
474
475## 0.2.50
476
477- New MCP "project" scope now allows you to add MCP servers to .mcp.json files and commit them to your repository
478
479## 0.2.49
480
481- Previous MCP server scopes have been renamed: previous "project" scope is now "local" and "global" scope is now "user"
482
483## 0.2.47
484
485- Press Tab to auto-complete file and folder names
486- Press Shift + Tab to toggle auto-accept for file edits
487- Automatic conversation compaction for infinite conversation length (toggle with /config)
488
489## 0.2.44
490
491- Ask Claude to make a plan with thinking mode: just say 'think' or 'think harder' or even 'ultrathink'
492
493## 0.2.41
494
495- MCP server startup timeout can now be configured via MCP_TIMEOUT environment variable
496- MCP server startup no longer blocks the app from starting up
497
498## 0.2.37
499
500- New /release-notes command lets you view release notes at any time
501- `claude config add/remove` commands now accept multiple values separated by commas or spaces
502
503## 0.2.36
504
505- Import MCP servers from Claude Desktop with `claude mcp add-from-claude-desktop`
506- Add MCP servers as JSON strings with `claude mcp add-json <n> <json>`
507
508## 0.2.34
509
510- Vim bindings for text input - enable with /vim or /config
511
512## 0.2.32
513
514- Interactive MCP setup wizard: Run "claude mcp add" to add MCP servers with a step-by-step interface
515- Fix for some PersistentShell issues
516
517## 0.2.31
518
519- Custom slash commands: Markdown files in .claude/commands/ directories now appear as custom slash commands to insert prompts into your conversation
520- MCP debug mode: Run with --mcp-debug flag to get more information about MCP server errors
521
522## 0.2.30
523
524- Added ANSI color theme for better terminal compatibility
525- Fixed issue where slash command arguments weren't being sent properly
526- (Mac-only) API keys are now stored in macOS Keychain
527
528## 0.2.26
529
530- New /approved-tools command for managing tool permissions
531- Word-level diff display for improved code readability
532- Fuzzy matching for slash commands
533
534## 0.2.21
535
536- Fuzzy matching for /commands