kimi-code/docs/en/reference/slash-commands.md

Slash Commands

Slash commands are built-in control commands provided by Kimi Code CLI in the interactive TUI, used to switch modes, manage sessions, view status, and more. Type / in the input box to trigger command completion; the candidate list filters in real time as you continue typing, and command aliases participate in matching as well.

After typing a full command name (such as /help), press Enter to execute it. If the /-prefixed input does not match any built-in or skill command, it is sent to the agent as an ordinary message.

::: tip Tip Some commands are only available in the idle state. Running them while the session is streaming a response or compacting the context will be blocked, with a hint to press Esc or Ctrl-C first to interrupt the current operation. The "Always available" column in the tables below marks commands that remain available during streaming or compacting. :::

Account and configuration

Command	Alias	Description	Always available
/login	—	Pick an account or platform and sign in: Kimi Code uses the OAuth device code flow, while the Moonshot AI Open Platform signs in with an API key.	No
/logout	—	Clear the credentials of the currently selected account (Kimi Code OAuth credentials, or the corresponding open platform provider config).	No
/connect [--refresh] [--url=<catalog-url>]	—	Configure a provider and model from a model catalog. The default catalog is bundled with the CLI; pass --refresh to fetch the latest catalog from models.dev, or --url to read it from a custom URL. See Providers and models — /connect and the model catalog.	No
/model	—	Switch the LLM model used by the current session.	No
/settings	/config	Open the settings panel inside the TUI.	Yes
/permission	—	Choose a permission mode.	Yes
/editor	—	Configure the external editor launched by Ctrl-G.	Yes
/theme	—	Switch the terminal UI color theme.	Yes

Session management

Command	Alias	Description	Always available
/new	/clear	Start a brand-new session, discarding the current context.	No
/sessions	/resume	Browse historical sessions and switch to or resume one.	No
/tasks	/task	Browse the background task list.	Yes
/fork	—	Fork a new session from the current one, preserving the full conversation history.	No
/title [<text>]	/rename	Without arguments, show the current session title; with an argument, set it as the new title (up to 200 characters).	Yes
/compact [<instruction>]	—	Compact the current conversation context to free up token usage; optionally pass a custom instruction telling the model what to preserve during compaction.	No
/init	—	Analyze the current codebase and generate AGENTS.md.	No
/export-md [<path>]	/export	Export the current session as a Markdown file. With no argument, writes to kimi-export-<short-id>-<timestamp>.md in the working directory; pass a path to choose the output location.	No
/export-debug-zip	—	Export the current session as a debug ZIP archive (mirrors kimi export). The archive always includes the active global diagnostic log.	No

Mode and runtime control

Command	Alias	Description	Always available
/yolo [on|off]	/yes	Toggle auto-approve mode. Without arguments, flip the current state; pass on/off explicitly to force the corresponding state. When enabled, ordinary tool call approvals are skipped; the Plan mode exit approval is not skipped.	Yes
/plan [on|off]	—	Toggle Plan mode. Without arguments, flip the current state; pass on/off explicitly to force the corresponding state. Toggling alone does not create an empty plan file.	Yes
/plan clear	—	Clear the current plan.	No

::: warning Note /yolo skips approval confirmation for ordinary tool calls. Make sure you understand the potential risks before enabling it. It does not skip the approval required to leave Plan mode; in Plan mode, Bash follows the same ordinary allow rules as /yolo. :::

Information and status

Command	Alias	Description	Always available
/help	/h, /?	Show keyboard shortcuts and all available commands.	Yes
/usage	—	Show token usage, context consumption, and quota information.	Yes
/status	—	Show the current session runtime status, including version, model, working directory, and permission mode.	Yes
/mcp	—	List the MCP servers in the current session and their connection status.	Yes
/plugins	—	Open the interactive plugin manager for user/global installs: install, inspect, enable, disable, confirm removal, reload, browse the official marketplace, and toggle plugin MCP servers. Shortcut subcommands remain available.	Yes
/version	—	Show the Kimi Code CLI version number.	Yes
/feedback	—	Submit feedback to help improve Kimi Code CLI.	Yes

Exit

Command	Alias	Description	Always available
/exit	/quit, /q	Exit Kimi Code CLI.	No

Dynamic skill commands

In addition to the built-in commands, user-activatable skills are automatically registered as slash commands under the skill: namespace:

/skill:<name> [extra text]

For example, /skill:code-style loads the content of the code-style skill and sends it to the agent; any text after the command is appended to the skill prompt, as in /skill:git-commits fix the login failure issue.

For convenience, skill commands also support a short form /<name> that omits the skill: prefix, provided the name is not already taken by a built-in command. In other words, /code-style falls back to matching /skill:code-style.

Kimi Code CLI ships with a built-in mcp-config skill for configuring MCP servers and handling MCP OAuth login. It still belongs to the skill namespace in completion and help (/skill:mcp-config), and it can also be invoked directly as /mcp-config.

Skill types that can be exposed as slash commands include prompt, inline, flow, and skills without an explicitly declared type. For skill installation and authoring, see Agent Skills.

::: info Note All skill commands are only available while the agent is idle; during streaming or compacting, press Esc or Ctrl-C first to interrupt the current operation. :::

::: info Note Flow-type skills are also exposed via /skill:<name>; there is no separate /flow: namespace. :::