Commands
Clew ships with 80+ built-in slash commands plus dynamically loaded skills and plugin commands. Commands can be local (terminal execution), prompt (text expanded to the model), or local-jsx (Ink UI panels).
Discovery
Type / in the REPL to see all available commands in your build. Commands are registered in src/commands.ts and loaded from built-in, plugin, skill, bundled, and workflow sources.
Session & Context
| Command | Description |
|---|
/clear | Clear conversation history and free up context |
/compact | Compress session context — log/trace snipping, consolidation, pruning |
/context | Visualize current context usage as a colored grid |
/resume | Resume a previous conversation |
/rename | Rename the current conversation |
/rewind | Restore code and/or conversation to a previous point |
/copy | Copy Claude's last response to clipboard (or /copy N for Nth-latest) |
/export | Export the current conversation to a file or clipboard |
/session | Show remote session URL and QR code |
/branch | Create a branch of the current conversation at this point |
/files | List all files currently in context |
Model & Provider
| Command | Description |
|---|
/model | Switch model or provider — shows all models grouped by provider (Enter=save default, s=session-only) |
/provider-select | Show or change the active AI provider and model |
/taste | Open taste preference-learning interactive menu (learn, forget, profile, events, decay, eval, on/off) |
/effort | Set effort level for model usage (low, medium, high, max) |
/cost | Show the total cost and duration of the current session |
/usage | Show plan usage limits |
/usage-credits | Configure usage credits to keep working when limits are hit |
/stats | Show your Clew usage statistics and activity |
/rate-limit-options | Show options when rate limit is reached |
/upgrade | Upgrade to Max for higher rate limits and more Opus |
Tools & Permissions
| Command | Description |
|---|
/tools | List all tools with usage stats and estimated token cost |
/permissions | Manage allow & deny tool permission rules |
/plan | Enable plan mode or view the current session plan |
/privacy-settings | View and update your privacy settings |
/datadog | Manage individual telemetry and analytics settings (Datadog, Anthropic, Global) |
/capabilities | Show available system capabilities (git, tmux, browser, network, shells) |
/hooks | View hook configurations for tool events (PreToolUse, PostToolUse, MessageDisplay, SessionStart) |
/sandbox-toggle | Toggle sandbox mode for tool execution |
Extensions (Plugins, MCP, Skills)
| Command | Description |
|---|
/plugin | Manage plugins — list, install, remove |
/plugin-details | Show plugin component inventory, hooks, MCP servers, and per-session token cost |
/reload-plugins | Activate pending plugin changes in the current session |
/mcp | Manage MCP servers (list shows pending approval for unapproved .mcp.json servers) |
/code-review | Review changed code for correctness bugs (low/medium/high effort; --fix applies changes, --comment posts PR comments) |
/simplify | Cleanup-only code review (reuse, efficiency, altitude) with auto-fix |
/skill | List available skills or show skill details |
/skills | List available skills |
Agents
| Command | Description |
|---|
/agent | Manage Clew AI Agents execution, state checkpoints, and approvals |
/agents | Open interactive agent dashboard with grouped sessions, PR columns, dispatch autocomplete, AI summaries, and keyboard shortcuts |
!bg <command> | Run a shell command as a persistent background agent task, visible in the agents dashboard with live status tracking |
/tasks | List and manage background tasks |
/daemon | Open an interactive 24/7 autonomous daemon control panel; subcommands still support start/stop/status/restart |
/task | Create scheduled tasks with an interactive form; subcommands still manage the autonomous task queue |
/loop | Run a prompt or slash command on a recurring interval (e.g. /loop 5m /check-deploy) |
/autofix-pr | Fix CI errors and address review comments on a PR using Clew on the web |
System & Diagnostics
| Command | Description |
|---|
/help | Show help and available commands |
/status | Show Clew status including version, model, account, API connectivity, and tool statuses |
/doctor | Diagnose and verify your Clew installation and settings |
/exit | Exit the REPL |
/version | Show version |
/release-notes | View release notes |
/upgrade | Upgrade to Max for higher rate limits |
Configuration
| Command | Description |
|---|
/config | Open config panel |
/theme | Change the theme |
/color | Set the prompt bar color for this session |
/keybindings | Open or create your keybindings configuration file |
/scroll-speed | Set mouse wheel scroll speed multiplier (1-20) |
/output-style | Deprecated: use /config to change output style |
/remote-env | Configure the default remote environment for teleport sessions |
/setting-sources | Configure setting sources |
/init | Run Setup hooks with init trigger |
Git & Code Review
| Command | Description |
|---|
/diff | View uncommitted changes and per-turn diffs |
/commit | Stage and commit changes |
/commit-push-pr | Commit, push, and create a PR |
/review | Review code changes |
/security-review | Run a security review |
/pr-comments | Get comments from a GitHub pull request |
/install-github-app | Set up Claude GitHub Actions for a repository |
Research & Search
| Command | Description |
|---|
/research | Deep source-grounded research across local files, wiki, and memory |
/memory | Edit Claude memory files |
/memory-search | Search memories using semantic search (cross-lingual) |
/searxng | Start, stop, or check status of the local SearXNG Docker container |
/explorer | Toggle file explorer sidebar |
Evaluation
| Command | Description |
|---|
/eval | Evaluate AI agent performance with the verification harness |
Utilities
| Command | Description |
|---|
/btw | Ask a quick side question without interrupting the main conversation |
/buddy | Configure your AI companion (Buddy) |
/desktop | Continue the current session in Claude Desktop |
/mobile | Show QR code to download the Claude mobile app |
/powerup | Interactive lessons teaching Clew features with animated demos |
/feedback | Submit feedback about Clew |
/onboarding | Run the interactive onboarding wizard |
/team-onboarding | Generate a teammate ramp-up guide from your local usage |
/add-dir | Add a new working directory |
/goal | Set a session goal |
/heapdump | Dump the JS heap to ~/Desktop |
/stickers | Order Clew stickers |
/thinkback | Your 2025 Clew Year in Review |
Feature-Gated Commands
These commands require specific environment variables to be enabled:
| Command | Flag | Description |
|---|
/bridge | BRIDGE_MODE=1 | Connect this terminal for remote-control sessions |
/voice | VOICE_MODE=1 | Toggle voice mode |
/brief | KAIROS=1 | Brief command |
/assistant | KAIROS=1 | Assistant mode |
/ant | USER_TYPE=ant | Toggle ant-only beta features |
/ultraplan | Internal | Extended plan mode |
/bughunter | Internal | Bug hunting workflow |
Interactive Autocomplete
Clew features an inline ghost text system that predicts the next argument as you type commands:
- Type
/goal followed by a space inline ghost text shows [condition] in dim text after the cursor
- Press Tab or → (Right Arrow) to accept the ghost text and fill the placeholder
- Supports progressive hints:
/goal fix [name] [priority] reveals remaining args as you fill each one
- Commands with
argumentHint (e.g. /plugin-details <plugin-name>) show static placeholder hints
For mid-input command completion, the same ghost text engine fills in partial command names:
- Type
/com ghost text shows mit for /commit
- Press Tab or → to complete the command name
Command Loading
Commands are loaded from multiple sources at startup:
- Built-in — Core commands from
src/commands/ (always available)
- Bundled skills — Skills shipped with the application
- Built-in plugin skills — From enabled built-in plugins
- Skill directory —
~/.claude/skills/ and project .claude/skills/
- Plugin commands — From installed plugins
- Plugin skills — From plugin skill directories
- MCP skills — From MCP servers (when
MCP_SKILLS feature is enabled)
- Dynamic skills — Discovered during file operations
Use /reload-plugins to reload plugin-provided commands without restarting. Remote-safe and bridge-safe commands have explicit allowlists in src/commands.ts.