Hooks & automation — page 3
Issues 401–600 of the 9,356 matching this topic. The trend chart and topic overview are on page 1.
- [BUG] Plugin cache keeps old versions and runs their hooks!
- [FEATURE] Add /skills command to list all available skills
- [BUG] No hook is working on Windows
- [BUG] Empty page after login in VSCode extension
- [Feature] Expose custom slash commands as tools
- [BUG] Desktop app blocks sending message when a custom slash command is combined with a URL
- Remote Control: built-in slash commands (/clear, /context) not intercepted — delivered to the model as plain text
- Fable 5: assistant text emitted before a tool call in the same turn is not shown to the user, but the model assumes it was
- [Windows] Visible console window flashes on every tool use in v2.1.143
- Docs: clarify whether /goal is supported in the Claude Code desktop app
- Rewind freezes terminal completely on macOS arm64 — must force-quit (v2.1.119)
- Claude Desktop: binaryName platform-switch causes macOS TCC prompt for 'Claude.exe'
- [FEATURE] Pre-exit Claude-action hook — let Claude perform actions before session ends
- Feature: Fine-grained compaction controls (partial compaction, boundary-aware, compaction context)
- Feature request: opt-in full bypassPermissions for protected directories (workaround included)
- Expose /rename as a tool so skills/commands can set session names programmatically
- [BUG] LSP tool is no longer being registered as a callable tool
- Add weekly and session usage percentages to status line JSON input
- [FEATURE] Automatic Co-authored-by attribution for Claude Code on the web
- [FEATURE] /alias command
- [BUG] SessionStart hooks never execute for local file-based marketplace plugins
- [BUG] Slash commands not loading in version 2.0.27 - "Unknown slash command" error
- Expose hooks configuration through the CLI and SDK
- [Feature Request] Configurable Timeout for Hooks
- Bug: Claude cannot execute custom slash commands from within conversations
- [Bug] Slash Command Execution Fails in Custom Project Scope
- Task/Todo tools (TaskCreate/TodoWrite) not exposed despite `todoFeatureEnabled: true` — regressed ~2026-07-21, appears account-gated
- Cross-session coordination for independently-launched Claude Code sessions
- /workflows slash command not recognized in VS Code extension
- VS Code extension ignores sandbox settings.json / /sandbox unavailable (approval-fatigue fix unreachable in IDE)
- [FEATURE] VS Code extension should honor remoteControlAtStartup (or expose claudeCode.remoteControlAtStartup)
- [FEATURE] Stable, documented schema for ~/.claude/projects/<encoded-cwd>/<uuid>.jsonl line types
- VS Code: [ Claude Extention ] Can't see what subagents are doing or if they're stuck.
- [FEATURE] Support Agent/Prompt Hooks Triggered Immediately After Claude Prompts for User Input
- Permission mode resets from 'Bypass permissions' to 'Edit automatically' mid-session
- [BUG] Emit OTEL tool_result event for slash command (skill) invocations
- [BUG] User-level custom commands in ~/.claude/commands/ not loaded into skill registry
- [BUG] Slash command autocomplete always selects wrong command regardless of typed input
- Add hook support for AskUserQuestion tool
- Feature Request: Expose conversation context metrics to status line hooks
- [FEATURE] Claude Code requires some built-in environment variables to facilitate the development of custom plugins.
- [Bug] Command autocomplete highlighting mismatches input pattern in slash commands
- [BUG] All hooks (SessionStart, Stop, UserPromptSubmit) require --debug flag to execute in v2.0.27
- [BUG] Notification hook only fires inconsistently (~25%) for permission prompts
- Unicode/UTF-8 Character Corruption in Hook System
- System-Level ExitPlanMode Tool Incorrectly Invoked During User-Defined /plan Command
- Feature Request: Invoke Subagents from Hooks for Deterministic Workflows
- Feature Request: Additional Claude Code Hook Events for Better Automation
- WorktreeRemove hook never fires when a worktree is removed at session exit
- Phantom user messages: text appears in transcript as user input that the user did not type
- [FEATURE] claude --usage
- [BUG]TUI banner duplication stacking repeatedly on macOS Terminal.app (v2.1.104) — blocking daily work
- Feature request: enable/disable individual skills via settings.json and session flags
- [FEATURE] Feature request: make /rename session name settable via --name flag or hooks
- Remote Control not enabled on Max plan account
- [BUG] - Bash tool commands extremely slow (~30s) despite being instant in normal terminal
- Hook scripts fail on Windows: backslash paths misinterpreted by Git Bash
- tmpclaude-*-cwd temp files not cleaned up on Windows after bash commands
- [FEATURE] Add agent identity to PreToolUse hook data
- [Bug] Hookify plugin import error: incorrect package namespace structure
- [Feature Request] Autonomous Messages After SessionStart Hook (startup/resume/clear)
- [BUG] Native installation / binary return API errors
- [BUG] MCP prompt not working with title
- [BUG] MCP slash commands unrecognized when followed by a prompt
- [BUG] Bash Permissions Not Working in Slash Command
- Worktree creation writes an ABSOLUTE core.hooksPath into config.worktree, so worktrees run the MAIN checkout's hooks
- [MODEL] Opus 4.8: runaway extended thinking (20k-64k output tokens/turn), replies to hallucinated user messages, fabricates forensic "evidence" when asked to investigate
- Add toggle to hide the branch status badge / PR widget
- Desktop sidebar ignores UserPromptSubmit hook `sessionTitle` output
- [BUG] Notification hooks not firing for permission_prompt in VS Code extension
- [FEATURE] Detect Jujutsu (.jj/) repositories in environment metadata
- Context is lost when Claude suggests restarting — should auto-save a session summary before restart
- Feature Request: Hook for assistant text output (PreResponse / PostResponse)
- [Feature Request] Add setting to skip protected directory prompts when bypassPermissions is enabled
- Allow tools/skills to programmatically clear context and inject a continuation prompt
- Stop hook with exit code 2 displays as 'Stop hook error' instead of feedback
- Feature request: Allow UserPromptSubmit hooks to modify prompt text
- Model ignores CLAUDE.md rules, makes unverified claims, reports false success
- [FEATURE] PostToolUse hooks: allow `updatedToolOutput` for built-in tools (context budget recovery)
- Expose rate limit utilization in statusLine JSON input
- Add /rename slash command support in VSCode extension
- autoCompact: false ignored - compaction triggers at ~78%
- [FEATURE] Sync Claude Code session name with VS Code terminal name
- [Ralph Wiggum Plugin] Windows 11 Compatibility Fixes for Stop Hook
- Feature Request: Allow Permission Mode Switching in PermissionRequest Hooks
- [Bug] Model Alias Not Supported in Custom Slash Command Configuration
- Incorrectly labels custom slash command in ~/.claude directory as '(project, gitignored)' instead of '(user)'
- [Feature Request] I want to be able to use multiple arguments with slash commands.
- Custom slash commands not loading after update to v1.0.62
- Max Account Token Incorrectly Generates Pro Account Credentials
- Feature Request: Auto-compact warning Hook
- [FEATURE] Let PostToolUseFailure hooks replace the error output the model sees (re-file of #77180)
- Session hangs mid-turn after a PreToolUse hook returns: tool never dispatches (2.1.220, Windows)
- Feature request: built-in, cross-platform way to list running Claude Code sessions and their state
- Permission-mode system prompt instructs the model to edit files via Bash, routing writes around PreToolUse hooks on Write|Edit|NotebookEdit
- bypass permissions mode steers agents off Read/Edit/Grep onto composed shell strings, reintroducing quoting and exit-code failures that the structured tools cannot have
- [BUG] LSP tool is pruned from all subagent tool sets in interactive sessions (present in the parent, and in subagents under -p)
- Built-in Task tools (TaskCreate/TaskGet/TaskList/TaskUpdate) intermittently unregistered, and can vanish mid-session — breaks the live task display
- [BUG] Invoking a skill (slash command) pins a large injected block and shrinks the assistant response into a tiny scroll area (VS Code extension)
- [FEATURE] Content-type color coding for assistant output (questions, commands, narrative)
- [FEATURE] Enable skill execution tracking in Cowork OTEL events
- WorktreeCreate hook causes indefinite hang regardless of hook command
- [FEATURE] Save conversation to local file (markdown/JSON export)
- PostToolUse hooks: support updatedBuiltinToolOutput for built-in tools
- Feature Request: Include session_name in hook input JSON
- Feature request: reliable interrupt/notification mechanism for inter-agent messaging
- [Feature Request] Auto-reconnect SSE MCP servers after server-side restart
- [FEATURE] Allow users to provide a reason/instruction when denying a tool call
- WorktreeRemove hook not invoked for subagent worktrees (isolation: worktree)
- Hookify warnings don't reach the model (only shown to user)
- Shell conflict with SCM Breeze causes '_safe_eval: command not found' error
- [BUG] PermissionRequest: Allow does not exit plan mode for `ExitPlanMode` tool.
- [BUG] Hook .sh scripts open in VSCode instead of executing on Windows Terminal/PowerShell
- [Windows] SessionStart hooks cause infinite hang during initialization
- Pipe character in regex patterns incorrectly triggers bash permission check
- Feature Request: Expose Active Permission Mode to Hooks and Statusline
- [BUG] SDK (TypeScript at least) fails to invoke hooks
- Claude Desktop blocks SendMessage entirely, which also removes subagent continuation
- Regression: slash command autocomplete no longer triggers mid-input (worked per v2.1.0, fixed once in v2.1.136)
- Prompt cache invalidated when PreToolUse/PostToolUse additionalContext changes during history rebuild
- [BUG] Repo-committed extraKnownMarketplaces + enabledPlugins never processed in claude.ai/code cloud sandboxes (hooks from the same settings.json work)
- [BUG] The Notification hook never fires when Claude Code runs as the VSCode native extension
- [FEATURE] Allow the Dynamic Workflows sandbox to execute external code / tools
- [FEATURE] Bandwidth-aware mode for metered connections (mobile hotspot, cellular tethering)
- [Feature Request] Allow programmatic/agent-initiated compaction
- [Feature Request] Do not include slash command output in conversation context (/release-notes takes 10% of context alone).
- System prompt size grew ~70K tokens between v2.1.89 and v2.1.96, making sessions unusable without frequent manual /compact
- [FEATURE] Add /lsp slash command for LSP server status
- Add 'compact context' option when accepting plans in Plan Mode
- Notification when Claude Code is waiting for user input
- [BUG] VSCode: ide_selection and ide_opened_file missing on first message when using a skill
- [BUG] Bash tool "avoid cd" instruction causes systematic failure in SSH remote commands
- [BUG] Docker-based MCP server containers not stopped when session ends
- [BUG] Claude (Opus 4.6) repeatedly executes actions out of order, ignoring a carefully designed workflow with checklist, hooks, skills, error logs, and explicit rules. D.B.C **3+ times in the same session**. Adding more rules does not fix the problem.
- Feature Request: Allow MCP tool access for Task subagents
- Plugin skills with user-invocable: true not appearing in slash command autocomplete
- Bug: /context command executes /compact instead
- Plugin hooks cannot execute scripts on Windows - /bin/bash cannot resolve file paths
- System prompt conflates skills and slash commands in <available_skills>
- [Bug] Opt+T keyboard shortcut conflicts with Turkish keyboard layout (₺ symbol)
- [Bug] Plugin hooks JSON output not captured
- Hooks broken again in v2.0.31 (regression after v2.0.30 fix)
- [Bug] Stop hooks with exit code 2 fail to continue when installed via plugins
- [BUG] Hooks stopped working in 2.0.27, work in 2.0.25
- [BUG] Is `/sandbox` command supported on WSL2? Getting "Unknown slash command: sandbox" error
- Agent keeps stopping all day i have to manually ask it why did it stop so it will carry on it's work even after it says it has to do more work
- Restore Blocking Stop Command Hooks in Claude Code
- [BUG] User settings hooks not loading in /hooks command (v1.0.51, v1.0.52)
- Custom slash commands docs are incorrect
- Mods off remotely on 2.1.288 despite docs saying on by default since 2.1.287 (rollout switch serves off)
- [BUG] VS Code extension: slash-command and `/goal` messages never collapse, so long /command arguments fill the chat panel
- [FEATURE] Detect Jujutsu (.jj/) repositories in environment metadata
- [Bug] `opusplan` silently falls back to Sonnet during plan mode (no signal)
- [BUG] UserPromptSubmit hook rejection reason never displayed (Desktop + VS Code)
- Working-state continuity: survive compaction and /clear (the long-session "goes dumb" problem)
- PreToolUse:Bash hooks not invoked in v2.1.176 (registered hooks silently inert)
- [FEATURE] Batch MCP reauthentication (`/mcp reauthenticate-all`)
- Workflow tool triggers on any message mentioning "workflow" — trigger is too broad
- [FEATURE] Selective session retention: allow pinning/bookmarking important sessions to prevent auto-cleanup deletion
- Cowork: Support Claude Code hooks (~/.claude/settings.json — UserPromptSubmit, Stop, etc.)
- [BUG] Claude Code on the web: declared plugins inactive on first session, require restart to fully load
- [BUG] /goal Stop hook emits markdown to stdout, fails JSON schema validation ("JSON validation failed")
- [BUG] CLI 2.1.145 sends literal "claude-opus-4-7[1m]" as model name on session resume, hits 404, silently falls back to 200K
- [FEATURE] Allow hooks (and/or skills) to set the chat session title
- [BUG] [BUG] SSH-Remote ccd-cli 2.1.128 still exits with code 1 on initialize with hookCallbackIds — regression of #50698
- Support slash command autocomplete when '/' is typed mid-prompt
- [BUG] PreToolUse hook "ask" permissionDecision is auto-approved in auto mode — no way to force prompt
- [BUG] Per-turn smoosh pipeline folds dynamic <system-reminder> text into tool_result.content, breaking prompt cache
- [Feature Request] Auto-enable remote control on session start
- RFC: Extensible Companion Framework — let the community build what /buddy started
- [Bug] v2.1.88 yanked release caused wasted session + significant token loss
- Korean (CJK) characters corrupted to U+FFFD in Claude Code responses on macOS + VSCode
- [FEATURE] Auto-adjust reasoning effort based on prompt complexity
- Security: Claude can rewrite its own hooks — Who watches the watchmen?
- Team agents fail to resolve 'inherit' model — stuck with invalid model ID
- Duplicate slash commands in git worktrees
- [Bug] Interactive input frozen on startup on Windows 11 PowerShell (v2.1.30–2.1.32)
- Add hook or customization point for Plan Mode exit/plan delivery
- Windows: Hook commands cause brief console window flash
- [BUG] `--json-schema` defined JSON doesn't exist in `structured_output`
- Feature Request: PostToolUseFailed hook for bash command failures
- Stop hook receives stale transcript - race condition between file write and hook invocation
- [DOCS] Missing PermissionRequest hook details in Hooks Guide and Input Reference schema
- [BUG] Prompt-based Stop hooks receive only metadata, not conversation content (regression confirmed in v2.0.42)
- Feature Request: Per-Command Thinking Mode Configuration
- Feature Request: Conversation Start Hook
- [QUESTION] Conda environment workflow guidance needed - inherited environment but missing shell functions
- use slash commands in print/headless/non-interactive mode
- [BUG] Desktop: first click on a plugin pane button only moves focus; the button needs a second click
- [FEATURE] Projects (beta): Support first-class Local Claude Code sessions as project threads
- [FEATURE] Desktop Code tab: allow custom display names for project groups in the sidebar
- [BUG] Slash command menu no longer triggers mid-input in the desktop app (regression; #44447 covered the terminal)
- [FEATURE] Expose the live subscription budget and per-task token accounting to the model, so it can plan a session instead of guessing
- [FEATURE] /nudge — send an ephemeral one-turn instruction that doesn't persist in conversation context
- [BUG] PermissionRequest hook 'allow' for ExitPlanMode is ignored since v2.1.199 — plan-approval chooser still blocks
- [MODEL] Self-report: six days of architectural drift on a customer project despite full hook + memory + skill enforcement
- [BUG] Skill subagent inherits parent's 1M-context tier but not extra-usage entitlement
- `# userEmail` and `# currentDate` auto-injected into MEMORY.md context cannot be suppressed
- Plugin autoUpdate doesn't update installed_plugins.json, leaving bundled hooks pinned to stale installPath
- Russian language (i18n) support for Claude Code UI