This feed tracks post-release changes to the Claude Code changelog. When Anthropic adds, modifies, or removes changelog entries after a version is released, those changes are detected and recorded here for transparency.
Added You should know, a built-in mod where a side agent watches your back and flags things you or Claude might miss. Turn it on with `/plugin enable cc-plugin-you-should-know@builtin` (for first-party sessions with telemetry on)
Detected at
Added You should know, an opt-in plugin where a side agent watches your back and flags things you or Claude might miss. Turn it on with `/plugin enable cc-plugin-you-should-know@builtin` (for first-party sessions with telemetry on)
Detected at
Changed AGENTS.md support to also work on Amazon Bedrock, Google Vertex AI, Microsoft Foundry, LLM gateways, and sessions with telemetry disabled
Detected at
Added AGENTS.md support: in a project with no CLAUDE.md, Claude Code reads AGENTS.md instead; change it under "Project instructions" in `/config`
Detected at
Added AGENTS.md support: in a project with no CLAUDE.md, Claude Code reads AGENTS.md instead; change it under "Project instructions" in `/config` (not yet on Bedrock, Vertex or Foundry)
Detected at
[VSCode] Fixed some claude.ai/code sessions opening in VS Code as an empty conversation with no messages
Detected at
Changed hook output over 10,000 characters to be saved to disk with a file path + a 2,000-character preview instead of being injected directly into context
Detected at
Changed hook output over 50K characters to be saved to disk with a file path + preview instead of being injected directly into context
Detected at
Fixed a permission you approved while a self-hosted runner was restarting being dropped when the session resumed, so the approved action now runs
Detected at
Fixed a SIGTERM arriving while a self-hosted runner was starting up leaving a stale active row until the lease expired; it now deregisters cleanly
Detected at
Added structured failure categories to self-hosted runner spawn and session failures, so hook errors, runner crashes and config errors can be told apart
Detected at
Self-hosted runner: added a `post-session` lifecycle hook that runs after the session ends and before the workspace is deleted, so you can snapshot uncommitted work or export logs; also made the child-process SIGTERM→SIGKILL window configurable (default unchanged at 5s)
Detected at
Added HTTP status and error text to `claude mcp list` and `/mcp` when a server fails to connect, and a warning for MCP config values with hidden leading or trailing whitespace
Detected at
Changed the Fable 5 usage-credits consent prompt to start with the decline option focused
Detected at
Added deprecation notification for npm installations - run `claude install` or see https://code.claude.com/docs/en/setup for more options
Detected at
Added deprecation notification for npm installations - run `claude install` or see https://docs.anthropic.com/en/docs/claude-code/getting-started for more options
Detected at
Subagents now run in the background by default, so Claude keeps working while they run and is notified when they finish (previously a gradual rollout)
Detected at
Agent teams: removed the `TeamCreate` and `TeamDelete` tools. With `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` set, every session now has one implicit team — spawn teammates directly with the Agent tool's `name` parameter, no setup step needed. The `team_name` parameter on the Agent tool is still accepted but ignored.
Detected at
Fixed Linux sandbox failing to start when `.claude/skills` or `.claude/hooks` is a symlink
Detected at
Self-hosted runner: added a `post-session` lifecycle hook that runs after the session ends and before the workspace is deleted, so you can snapshot uncommitted work or export logs; also made the child-process SIGTERM→SIGKILL window configurable (default unchanged at 5s)
Detected at
Hardened the REPL tool sandbox against prototype-pollution and thenable-based escapes
Detected at
Hardened the REPL tool sandbox against prototype-pollution and thenable-based escapes
Detected at
Added the `Workflow` tool for deterministic multi-agent orchestration. It is off by default — set `CLAUDE_CODE_WORKFLOWS=1` to enable
Detected at
Hardened REPL and Workflow tool sandboxes against prototype-pollution and thenable-based escapes
Detected at
Fixed `! <cmd>` exec sessions not responding to Ctrl+C while attached — now interrupts the running command
Detected at
Fixed agent view shell-command rows lingering under Working after completion, and pressing Enter on a completed row re-running the command after its output expired
Detected at
Fixed `/remote-control` retries appearing stuck on "connecting…" — each retry now shows its result
Detected at
Fixed blank remote-session transcript when certain messaging tools are unavailable
Detected at
Fixed `/remote-control` retries appearing stuck on "connecting…" — each retry now shows its result, and unenrolled trusted-device failures are caught up-front
Detected at
Claude Code and installer now use `https://downloads.claude.ai/claude-code-releases` instead of `https://storage.googleapis.com/claude-code-dist-86c565f3-f756-42ad-8dfa-d59b1c096819/claude-code-releases`
Detected at
Added push notification tool — Claude can send mobile push notifications when Remote Control and "Push when Claude decides" config are enabled
Detected at
Added `CLAUDE_CODE_DISABLE_NONSTREAMING_FALLBACK` env var to disable the non-streaming fallback when streaming fails
Detected at
Plugin options (`manifest.userConfig`) now available externally — plugins can prompt for configuration at enable time, with `sensitive: true` values stored in keychain (macOS) or protected credentials file (other platforms)
Detected at
Claude can now reference the on-disk path of clipboard-pasted images for file operations
Detected at
`Ctrl+L` now clears the screen and forces a full redraw — use this to recover when Cmd+K leaves the UI partially blank. Use `Ctrl+U` or double-Esc to clear prompt input.
Detected at
`--bare -p` (SDK pattern) is ~14% faster to the API request
Detected at
Memory: `MEMORY.md` index now truncates at 25KB as well as 200 lines
Detected at
Disabled `AskUserQuestion` and plan-mode tools when `--channels` is active
Detected at
Fixed API 400 error when a pasted image was queued during a failing tool call
Detected at
Fixed MCP tool calls hanging indefinitely when an SSE connection drops mid-call and exhausts its reconnection attempts
Detected at
Fixed Remote Control session titles showing raw XML when a background agent completed before the first user message
Detected at
Fixed remote sessions forgetting conversation history after a container restart due to progress-message gaps in the resumed transcript chain
Detected at
Fixed remote sessions requiring re-login on transient auth errors instead of retrying automatically
Detected at
Fixed `rg ... | wc -l` and similar piped commands hanging and returning `0` in sandbox mode on Linux
Detected at
Fixed voice input hold-to-talk not activating when a CJK IME inserts a full-width space
Detected at
Fixed `--worktree` hanging silently when the worktree name contained a forward slash
Detected at
[VSCode] Added Esc-twice (or `/rewind`) to open a keyboard-navigable rewind picker
Detected at
[VSCode] Fixed "Fork conversation from here" and rewind actions failing silently after the session cache goes stale
Detected at
Fixed tool search to activate even with `ANTHROPIC_BASE_URL` as long as `ENABLE_TOOL_SEARCH` is set.
Detected at
Changed tool search to bypass the third-party proxy gate when the environment variable is set (replaces `CLAUDE_CODE_PROXY_SUPPORTS_TOOL_REFERENCE`, now removed)
Detected at
Added `claude plugins` as an alias for `claude plugin`
Detected at
Improved `claude plugin validate` to explain that marketplace.json source paths are relative to the repo root when rejecting `../` paths
Detected at
Reduced false-positive bash permission prompts — tree-sitter parsing now handles `find -exec`, variable assignments, command substitutions, and many other patterns that previously triggered unnecessary prompts. Also fixed tree-sitter not loading in npm-installed versions
Detected at
Fixed config file corruption that could wipe authentication when multiple Claude Code instances ran simultaneously
Detected at
Managed settings can now be set via macOS plist or Windows Registry. Learn more at https://code.claude.com/docs/en/settings#settings-files
Detected at
Improved MCP OAuth authentication with step-up auth support and discovery caching, reducing redundant network requests during server connections
Detected at
Added `--worktree` (`-w`) flag to start Claude in an isolated git worktree
Detected at
Subagents support `isolation: "worktree"` for working in a temporary git worktree
Detected at
Added Ctrl+F keybinding to kill background agents (two-press confirmation)
Detected at
Agent definitions support `background: true` to always run as a background task
Detected at
Plugins can ship `settings.json` for default configuration
Detected at
Fixed file-not-found errors to suggest corrected paths when the model drops the repo folder
Detected at
Fixed ENAMETOOLONG errors for deeply-nested directory paths
Detected at
Improved startup performance by deferring Zod schema construction
Detected at
Improved prompt cache hit rates by moving date out of system prompt
Detected at
Added one-time Opus 4.6 effort callout for eligible users
Detected at
Fixed image dimension limit errors to suggest /compact
Detected at
Fixed Opus 4.6 launch announcement showing for Bedrock/Vertex/Foundry users
Detected at
Improved error message for many-image dimension limit errors with /compact suggestion
Detected at
Added guard against launching Claude Code inside another Claude Code session
Detected at
Fixed Agent Teams using wrong model identifier for Bedrock, Vertex, and Foundry customers
Detected at
Fixed a crash when MCP tools return image content during streaming
Detected at
Fixed /resume session previews showing raw XML tags instead of readable command names
Detected at
Improved model error messages for Bedrock/Vertex/Foundry users with fallback suggestions
Detected at
Fixed plugin browse showing misleading "Space to Toggle" hint for already-installed plugins
Detected at
Fixed hook blocking errors (exit code 2) not showing stderr to the user
Detected at
Added `speed` attribute to OTel events and trace spans for fast mode visibility
Detected at
Fixed AWS auth refresh hanging indefinitely by adding a 3-minute timeout
Detected at
Added guard against launching Claude Code inside another Claude Code session
Detected at
Fixed Agent Teams using wrong model identifier for Bedrock, Vertex, and Foundry customers
Detected at
Fixed /resume session previews showing raw XML tags instead of readable command names
Detected at
Fixed a crash when MCP tools return image content during streaming
Detected at
Improved model error messages for Bedrock/Vertex/Foundry users with fallback suggestions
Detected at
Added `speed` attribute to OTel events and trace spans for fast mode visibility
Detected at
Fixed hook blocking errors (exit code 2) not showing stderr to the user
Detected at
Fixed plugin browse showing misleading "Space to Toggle" hint for already-installed plugins
Detected at
Fixed /resume showing interrupt messages as session titles
Detected at
Fixed Opus 4.6 launch announcement showing for Bedrock/Vertex/Foundry users
Detected at
Improved error message for many-image dimension limit errors with /compact suggestion
Detected at
Fixed structured-outputs beta header being sent unconditionally on Vertex/Bedrock
Detected at
Fixed spurious warnings for non-agent markdown files in `.claude/agents/` directory
Detected at
Fixed spurious warnings for non-agent markdown files in `.claude/agents/` directory
Detected at
Fixed structured-outputs beta header being sent unconditionally on Vertex/Bedrock
Detected at
Improved error message for many-image dimension limit errors with /compact suggestion
Detected at
Fixed Opus 4.6 launch announcement showing for Bedrock/Vertex/Foundry users
Detected at
Fixed /resume showing interrupt messages as session titles
Detected at
Fixed plugin browse showing misleading "Space to Toggle" hint for already-installed plugins
Detected at
Fixed hook blocking errors (exit code 2) not showing stderr to the user
Detected at
Added `speed` attribute to OTel events and trace spans for fast mode visibility
Detected at
Fixed a crash when MCP tools return image content during streaming
Detected at
Improved model error messages for Bedrock/Vertex/Foundry users with fallback suggestions
Detected at
Fixed /resume session previews showing raw XML tags instead of readable command names
Detected at
Fixed Agent Teams using wrong model identifier for Bedrock, Vertex, and Foundry customers
Detected at
Added guard against launching Claude Code inside another Claude Code session
Detected at
Permissions now respect content-level `ask` over tool-level `allow`. Previously `allow: ["Bash"], ask: ["Bash(rm *)"]` allowed all bash commands, but will now permission prompt for `rm`.
Detected at
VSCode: Enabled Claude in Chrome integration
Detected at
Added customizable keyboard shortcuts via `~/.claude/keybindings.json`. Run `/keybindings` to get started. Learn more at https://code.claude.com/docs/en/keybindings
Detected at
Added customizable keyboard shortcuts via `~/.claude/keybindings.json`. Run `/keybindings` to get started. Learn more at https://code.claude.com/docs/en/keybindings
Detected at
Added `IS_DEMO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
Detected at
Added `CLAUDE_CODE_HIDE_ACCOUNT_INFO` environment variable to hide email and organization from the UI, useful for streaming or recording sessions
Detected at
Added `plan_mode_required` spawn parameter for teammates to require plan approval before implementing changes
Detected at
Claude now suggests prompts to speed up your workflow: press Tab to accept or Enter to submit
Detected at
Bedrock: Add support for `aws login` AWS Management Console credentials
Detected at
Unshipped AgentOutputTool and BashOutputTool, in favor of a new unified TaskOutputTool
Detected at
Added support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
Detected at
Add support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
Detected at
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
Detected at
Removed legacy SDK entrypoint. Please migrate to @anthropic-ai/claude-agent-sdk for future SDK updates: https://docs.claude.com/en/docs/claude-code/sdk/migration-guide
Detected at
Plugin documentation available at https://code.claude.com/docs/en/plugins
Detected at
Plugin documentation available at https://docs.claude.com/en/docs/claude-code/plugins
Detected at
Added SlashCommand tool, which enables Claude to invoke your slash commands. https://code.claude.com/docs/en/slash-commands#SlashCommand-tool
Detected at
Added SlashCommand tool, which enables Claude to invoke your slash commands. https://docs.claude.com/en/docs/claude-code/slash-commands#SlashCommand-tool
Detected at
Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://code.claude.com/docs/en/output-styles
Detected at
Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
Detected at
Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://code.claude.com/docs/en/hooks
Detected at
Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://docs.claude.com/en/docs/claude-code/hooks
Detected at
Fix issue with excessive iTerm notifications
Detected at
VSCode Extension: Added support for VSCode's secondary sidebar (VSCode 1.97+), allowing Claude Code to be displayed on the right sidebar while keeping the file explorer on the left. Requires setting "sidebar" as "Preferred Location" in the config.
Detected at
VSCode Extension: Added "Preferred Location" setting to configure where Claude opens by default (sidebar or panel)
Detected at
VSCode Extension: Added keyboard shortcut (Cmd+N / Ctrl+N) to start new conversations
Detected at
Add support for Microsoft Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
Detected at
Add support for Azure AI Foundry! See https://code.claude.com/docs/en/azure-ai-foundry
Detected at
Fixed messages returning null `stop_reason` and `stop_sequence` values
Detected at
Slash Commmands: Added CLAUDE_PROJECT_DIR and CLAUDE_PLUGIN_ROOT env vars to bash command processing
Detected at
Deprecated output styles. Review options in `/output-style` and use --system-prompt-file, --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
Detected at
Deprecated output styles. Review options in `/output-style` and use --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
Detected at
Fixed conversation compaction to respect existing compact boundaries, preventing duplicate summarization of already-compacted messages
Detected at
Deprecated output styles. Review options in `/output-style` and use --system-prompt, --append-system-prompt, CLAUDE.md, or plugins instead
Detected at
Deprecated output styles. Review options in `/output-style` and use CLAUDE.md or plugins instead
Detected at
Edit the plan in your text editor using CTRL+G
Detected at
Support MCP `structuredContent` field in tool responses
Detected at
Added an interactive question tool
Detected at
Claude will now ask you questions more often in plan mode
Detected at
Added Haiku 4.5 as a model option for Pro users
Detected at
Fixed an issue where queued commands don't have access to previous messages' output
Detected at
Fixed bash permission matching for commands with environment variables
Detected at
Improved VSCode extension command registry and sessions dialog user experience
Detected at
Enhanced sessions dialog responsiveness and visual feedback
Detected at
Fixed IDE compatibility issue by removing worktree support check
Detected at
Fixed security vulnerability where Bash tool permission checks could be bypassed using prefix matching
Detected at
Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.claude.com/en/docs/claude-code/output-styles
Detected at
Released output styles, including new built-in educational output styles "Explanatory" and "Learning". Docs: https://docs.anthropic.com/en/docs/claude-code/output-styles
Detected at
Released hooks. Special thanks to community input in https://github.com/anthropics/claude-code/issues/712. Docs: https://docs.claude.com/en/docs/claude-code/hooks
Detected at
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
Detected at
/model now validates provided model names
Detected at
Support NO_PROXY environment variable to bypass proxy for specified hostnames and IPs
Detected at
Hooks: Introduced SessionEnd hook
Detected at
@-mention: Add ~/.claude/\* files to suggestions for easier agent, output style, and slash command editing
Detected at
@-mention: Add ~/.claude/* files to suggestions for easier agent, output style, and slash command editing
Detected at
Customizable status line: add your terminal prompt to Claude Code with /statusline
Detected at
Transcript mode (Ctrl+R): Changed Esc to exit transcript mode rather than interrupt
Detected at
Settings: Added `--settings` flag to load settings from a JSON file
Detected at
Settings: Fixed resolution of settings files paths that are symlinks
Detected at
OTEL: Fixed reporting of wrong organization after authentication changes
Detected at
Slash commands: Fixed permissions checking for allowed-tools with Bash
Detected at
IDE: Added support for pasting images in VSCode MacOS using ⌘+V
Detected at
IDE: Added `CLAUDE_CODE_AUTO_CONNECT_IDE=false` for disabling IDE auto-connection
Detected at
Added `CLAUDE_CODE_SHELL_PREFIX` for wrapping Claude and user-provided shell commands run by Claude Code
Detected at
Added --settings flag to load settings from a JSON file
Detected at
Rename CLAUDE_CODE_BASH_PREFIX to CLAUDE_CODE_SHELL_PREFIX
Detected at
Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/commands/frontend/component.md is now /frontend:component, not /component.
Detected at
Custom slash commands: Restored namespacing in command names based on subdirectories. For example, .claude/frontend/component.md is now /frontend:component, not /component.
Detected at
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
Detected at
Released [hooks](https://docs.anthropic.com/en/docs/claude-code/hooks). Special thanks to community input in [Github Issues](https://github.com/anthropics/claude-code/issues/712)
Detected at
/resume slash command to switch conversations within Claude Code
Detected at
Breaking change: --print JSON output now returns nested message objects, for forwards-compatibility as we introduce new metadata fields
Detected at
Breaking change: --print JSON output now returns nested message objects, for forwards-compatability as we introduce new metadata fields
Detected at
Renamed tools for consistency: LSTool -> LS, View -> Read, etc.
Detected at
Renamed tools for consistency: AgentTool -> Task, LSTool -> LS, View -> Read, etc.
Detected at
Added support for refreshing dynamically generated API keys (via apiKeyHelper), with a 5 minute TTL
Detected at
Add support for refreshing dynamically generated API keys (via `apiKeyHelper`), with a 5 minute TTL
Detected at
Updated spinner to indicate tokens loaded and tool usage
Detected at
Update spinner to indicate tokens loaded and tool usage
Detected at
Improved performance for filename auto-complete
Detected at
Run one-off MCP servers with `claude --mcp-config <path-to-file>`
Detected at