Why this guide
Most people skim /help once and end up reaching for the same five or six commands every day. There’s a long tail of built-ins that quietly stay buried even though they need no setup and a single line to run. A handful of them change how a daily session feels.
Command descriptions follow the official Commands reference as of October 2026 (v2.1.314).
When you need a quick break
-
/radioOpen Claude FM lo-fi radio in your browser. Prints the stream URL when no browser is available. Since 2.1.251 it’s also available on Bedrock, Vertex AI, Foundry, and Claude Platform on AWS, and when telemetry is disabled.

-
/btw [question]Ask a side question about the current session without adding to the conversation. Useful when you want to confirm something without polluting context. Run
/btwwithout a question to browse your most recent side question and earlier answers. -
/color [color|default]Set the prompt bar color for the current session (
red,blue,green,yellow,purple,orange,pink,cyan). Usedefaultto reset, or run with no argument to pick a random color. Helpful for telling parallel sessions apart at a glance.
Move sessions freely
-
/teleport(alias/tp)Pull a cloud session (Claude Code on the web) into this terminal — opens a picker, then fetches the branch and conversation. Requires a claude.ai subscription.
-
/remote-control(alias/rc)The inverse: make this session available for Remote Control from claude.ai. Useful for checking in or directing work from a phone.
-
/branch [name]Create a branch of the current conversation at this point, so you can try a different direction without losing the conversation as it stands. Switches you into the branch and preserves the original, which you can return to with
/resume. -
/fork [prompt]Copy the current conversation into a new background session and keep working here. Pass a prompt and the copy starts working on it immediately.
/forkbecame its own command in 2.1.212 (it used to be an alias of/branch), and since 2.1.221 the copy creates a worktree of its own before making code changes. To hand a side task to a subagent that reports back into this conversation, use/subtask. -
/clear [name](aliases/reset,/new)Start a new conversation with empty context. Pass a name to label the previous conversation in the
/resumepicker.
When context feels heavy
-
/context [all]Visualize current context usage as a colored grid. Shows optimization suggestions for context-heavy tools, memory bloat, and capacity warnings. In fullscreen mode,
/contextcollapses the per-item breakdown; passallto expand it. -
/compact [instructions]Free up context by summarizing the conversation so far. Pass focus instructions to steer how the summary is shaped.
Get results out
-
/copy [N]Copy the last assistant response (or the Nth-latest with
/copy 2) to clipboard. When code blocks are present, opens an interactive picker to select individual blocks or the full response. Presswin the picker to write the selection to a file instead — useful over SSH. -
/export [filename]Export the current conversation as plain text. With a filename, writes directly to that file; without, opens a dialog to copy to clipboard or save to a file.
-
/diffReview the changes in your working tree, including the edits Claude has made so far. With fullscreen rendering, the diff panel opens beside the conversation and keeps updating while you work (requires v2.1.287 or later, a git repository, and a terminal at least 110 columns wide). With the classic renderer, the diff dialog opens above the prompt: Up/Down to select a file, Enter to open its diff, and Tab to move to the
sourcepicker and switch betweenCurrent(the working tree) and per-turn views (T1,T2, …). Great for a final pass before opening a PR.
Review workflows
-
/simplify [target]Bundled skill. Reviews the changed code for cleanup opportunities and applies the fixes. Four review agents run in parallel, covering reuse of existing helpers, simplification, efficiency, and whether the change is at the right level of abstraction (cleanup-only since 2.1.154). It doesn’t look for correctness bugs — use
/code-reviewfor that. Pass a path or PR reference to review a specific target. -
/review [level] [pr#|branch|path]Since 2.1.223, an alias of
/code-review: reviews the current diff, or a PR number, branch, or path you pass, for correctness bugs, and takes the same effort levels (lowthroughmax) and flags such as--fixand--comment. Example:/review 1234. For a deep cloud review, use/code-review ultra;/ultrareviewis an alias. -
/security-reviewAnalyze the changes on your current branch for security vulnerabilities. Reviews the diff between your branch and origin’s default branch, identifying risks like injection, auth issues, and data exposure. Needs an
originremote.
Automate sessions
-
/loop [interval] [prompt](alias/proactive)Bundled skill. Run a prompt repeatedly while the session stays open. Omit the interval and Claude self-paces between iterations. Omit the prompt and Claude runs the built-in maintenance prompt or your
loop.md. Example:/loop 5m check if the deploy finished. -
/effort [level|auto|status]Set the effort level immediately:
lowtoxhigh,max, orauto;statusprints it.maxis session-only. With no argument, opens an interactive slider — arrow keys to pick, Enter to apply. -
/advisor [model|off]Enable the advisor tool, which consults a second model for guidance at key moments — before committing to an approach, when an error keeps recurring, or before declaring a task done. For example, Sonnet main + Opus advisor keeps routine work fast and escalates planning, ambiguous failures, and completion checks to Opus. Accepts
fable,opus,sonnet, or a full model ID; without an argument, opens a picker. Experimental and requires the Anthropic API (not available on Amazon Bedrock, Claude Platform on AWS, Google Cloud, or Microsoft Foundry). Turn it off with/advisor off. Docs. -
/plan [description]Enter plan mode directly from the prompt. Pass an optional description to enter plan mode and immediately start with that task — for example
/plan fix the auth bug.
Wrap up
Type / inside a Claude Code session to see the full menu. The list looks intimidating, but you really only end up using five to ten of them regularly — slip two or three of these into your habit and the daily flow gets noticeably lighter.