Command Reference

Core Claude Commands (4)

Command	Description
/claude	Send a prompt to Claude Code. Supports prompt and session_id options.
/claude-thread	Start a new Claude conversation in a dedicated Discord thread. Options: name (thread title), prompt.
/resume	Resume (continue) a previous Claude conversation with an optional follow-up prompt.
/claude-cancel	Cancel the currently running Claude operation.
/fast	Toggle Opus 4.6 fast mode — 2.5x faster responses, same quality.

Enhanced Claude Commands (4)

Command	Description
/claude-enhanced	Enhanced Claude interaction with model options and streaming.
/claude-models	List all available Claude models (dynamically fetched).
/claude-sessions	Manage and view Claude conversation sessions.
/claude-context	Set additional context for Claude interactions.

Development Tools (7)

Command	Description
/claude-explain	Explain code or concepts. Options: content, detail_level, include_examples.
/claude-debug	Debug errors or code. Options: error_or_code, language, context_files.
/claude-optimize	Optimize code. Options: code, focus (speed/memory/readability/all), preserve_functionality.
/claude-review	Code review. Options: code_or_file, review_type, include_security, include_performance.
/claude-generate	Generate code. Options: request, type (function/class/module/test/api/config), style.
/claude-refactor	Refactor code. Options: code, goal, preserve_behavior, add_tests.
/claude-learn	Learn a topic. Options: topic, level (beginner/intermediate/advanced/expert), include_exercises, step_by_step.

Info & Control Commands (3)

Command	Description
/claude-info	View account info, available models, and MCP server status. Options: section (all/account/models/mcp). Works with or without an active session.
/rewind	Rewind file changes to a specific conversation turn. Options: turn (number), dry_run (preview changes without applying).
/claude-control	Mid-session controls. Options: action (interrupt/set-model/set-permissions/stop-task/status), value. Change model, permissions, or stop background tasks without restarting. In default/acceptEdits mode, unapproved tools trigger interactive Allow/Deny buttons.

Settings Commands (2)

Command	Description
/settings	Unified settings hub. Categories: show, bot, claude, modes, output, proxy, developer, reset. Action and value fields have autocomplete.
/quick-model	Quickly switch the active Claude model.

Task & Agent Management (3)

Command	Description
/todos	Task management. Actions: list, add, complete, generate, prioritize. Priority levels: low/medium/high/critical. Persists to disk.
/mcp	MCP (Model Context Protocol) server management. Actions: list, add, remove, test, status, toggle, reconnect. Reads from .claude/mcp.json. Toggle/reconnect work mid-session.
/agent	Run specialized AI agents. Actions: list, start, stop, status. 7 built-in agents.

Git Operations (6)

Command	Description
/git	Run git commands from Discord.
/worktree	Create a new git worktree with its own Discord channel.
/worktree-list	List all active worktrees.
/worktree-remove	Remove a worktree and its channel.
/worktree-bots	Manage bot instances per worktree.
/worktree-kill	Kill a worktree bot instance.

Shell Management (4)

Command	Description
/shell	Execute shell commands on the host.
/shell-input	Send input to a running shell process.
/shell-list	List active shell processes.
/shell-kill	Kill a running shell process.

System Monitoring (11)

Command	Description
/system-info	System information overview.
/processes	List running processes.
/system-resources	CPU, memory, and disk usage.
/network-info	Network configuration and connections.
/disk-usage	Detailed disk usage analysis.
/env-vars	View environment variables (filtered for safety).
/system-logs	View system logs.
/port-scan	Scan for open ports.
/service-status	Check service status.
/uptime	System uptime.
/screenshot	Capture a screenshot of the host display. Options: delay.

Utilities (4)

Command	Description
/status	Bot status and health check.
/pwd	Show current working directory.
/shutdown	Gracefully stop the bot.
/help	Show all available commands.