Interactive sessions

The Cluster Build TUI — what the assistant can do, how approvals and plan mode work, keybindings, and the full slash-command reference.

Interactive sessions

Run cluster with no arguments to open an interactive session in your current directory:

cluster

You type a request, the assistant works — reading files, running commands, and proposing edits — and you stay in the loop, approving changes as they come. The interactive UI needs a real terminal (TTY); for scripts and pipelines use cluster exec instead.

What the assistant can do

The model has a built-in toolset scoped to your current directory:

ToolWhat it does
shellRun shell commands, including long-running background processes
read_fileRead a file by line range, byte range, or surrounding code block
list_dir / findList directories and find files (gitignore-aware)
grepSearch file contents with ripgrep
web_search / web_fetchSearch the web and fetch page content through Cluster's search service (read-only, auto-approved)
apply_patchApply edits to files
request_user_inputAsk you 1–3 short multiple-choice questions when a decision can't be settled from the repo (available in plan mode; enable in default mode with request_user_input_default_mode)

grep requires ripgrep (rg) on your PATH. Background commands started with shell are polled and stopped by the assistant automatically.

This is the core set. Run /tools in a session to see the complete, live list — it also includes background-shell helpers and any tools added by your skills, MCP servers, or cloud commands.

Approvals

Before any action that changes your files or system, the assistant pauses and asks for approval. Read-only actions — reading files, listing directories, searching — run without prompting.

Two flags change this behavior at startup (both are global, so they also apply to cluster exec):

  • cluster --planplan mode (see below).
  • cluster --dangerously-skip-approvals (alias cluster --yolo) — auto-approve every tool call.

--dangerously-skip-approvals / --yolo lets the assistant edit files and run commands with no confirmation. Use it only in throwaway or sandboxed environments.

Plan mode

Plan mode blocks mutating tools for the session — the model can still read and search, but any edit or command is denied instead of run, so you get a plan before anything touches your project. Turn it on at startup with cluster --plan, toggle it mid-session with /plan or Shift+Tab, and note that it cannot be overridden by --dangerously-skip-approvals.

In plan mode the assistant can also call request_user_input to ask you a few structured multiple-choice questions when a choice materially changes the plan.

Keybindings

KeyAction
EnterSend your message
Ctrl-CStop the current turn
Ctrl-DQuit (when the input is empty)
Ctrl-TOpen the transcript pager
Shift+TabToggle plan mode

Slash commands

Inside a session, type / to see available commands. The full set:

CommandAction
/help (/h)List all commands
/model <id>Switch model
/effort <level>Set reasoning effort (minimalmax)
/planToggle read-only plan mode
/toolsList the assistant's available tools
/diffShow the working-tree git diff
/statusSession summary — model and token usage
/compactSummarize older context to free up room
/resumeBrowse and switch sessions
/clear (/reset)Start a fresh conversation
/stopStop background shell processes
/theme <light|dark>Switch the color theme
/skillsBrowse your custom skills
/mcpInspect and manage MCP servers — reconnect, enable/disable, and authenticate without restarting
/goal <text>Set a persistent goal the assistant works toward
/loop <prompt>Schedule a recurring run
/ultraplan <request>Plan against your repos on a cloud VM
/ultrareviewLaunch a cloud multi-agent code review
/quit (/q)Exit

Next steps

On this page