Skip to main content
Extending Qoder CLI

Built-in Capabilities Overview

An overview of Qoder CLI CN's built-in slash commands, Subagents, Skills, and workflows

Qoder CLI CN ships with a full set of ready-to-use capabilities out of the box, no extra configuration required. These built-in capabilities fall into four categories:
  • Slash Commands: commands typed with a leading / in the interactive interface, for managing sessions, configuration, extensions, and workflows.
  • Built-in Subagents: dedicated subagents preloaded with Qoder CLI CN, handling specific tasks such as code exploration, solution design, memory management, status line configuration, and usage Q&A.
  • Built-in Skills: preloaded reusable skills that package common workflows (review, debugging, security review, looped execution, and more) into directly invocable capabilities.
  • Workflows: orchestrate multi-step tasks into repeatable processes.
This page gives an overview of and navigation across these four categories. For per-item details such as parameters, aliases, and scopes, see Built-in Agents and Skills and Slash Commands.

Built-in slash commands

In the interactive interface, typing / pops up the command list with fuzzy matching. Slash commands let you perform management and auxiliary operations directly within the session without leaving the conversation. By purpose they roughly group as follows.

Session management

CommandPurpose
/newStart a new empty session.
/clearClear the screen and conversation history, starting a new session (similar to /new).
/continueContinue the previous session.
/resumePick and resume a session from the history list.
/renameRename the current session.
/rewindRewind to an earlier checkpoint, restoring code or conversation state.
/exportExport the current session content.
/compactCompress the current conversation context to free context window space.
/quitExit Qoder CLI CN.
For the full description of session management, see Managing Sessions and Undo and Restore.

Configuration and account

CommandPurpose
/loginSign in and complete authentication.
/logoutSign out of the current account.
/modelSelect or switch the model.
/themeSwitch the interface theme.
/editorConfigure the external editor.
/settingsView and adjust settings.
/initInitialize project configuration (generate the project instructions file, etc.).
/profileView the user profile.
/privacyView and adjust privacy-related settings.
/permissionsManage permissions (view and adjust permission rules).
/vimToggle Vim editing mode.
/voiceEnable voice input.
/statuslineConfigure a custom status line.
For configuration layers and precedence, see Settings Files and Precedence.

Extensions and tools

CommandPurpose
/agentsManage Subagents (create, edit, view).
/skillsManage Skills.
/mcpManage configured MCP servers.
/hooksManage Hooks.
/pluginsManage plugins.
/marketplaceBrowse and install extensions.
/toolsView the available built-in tools.
For detailed usage of each extension mechanism, see the sections below and MCP Reference, Hooks Reference, and Plugins Reference.

Workflows and tasks

CommandPurpose
/planEnter Plan (plan before executing) mode.
/tasksManage tasks.
/workflowsManage workflows.
/kanbanView tasks in a kanban view.
/reviewReview code changes.
/diffView diffs.

Information and usage

CommandPurpose
/helpView help.
/docsOpen the documentation.
/shortcutsView keyboard shortcuts.
/usageCheck usage and quota.
/insightsView session and usage insights.
/statusView the current status.
/contextView the context composition.
/release-notesView release notes.
/feedbackSubmit feedback.
/aboutAbout Qoder CLI CN.

Command availability

Some built-in commands are shown or hidden dynamically based on the current environment and settings:
  • Feature-gated: commands such as Subagent, Plan mode, Skills, workflows, and the extensions marketplace only appear when the corresponding capability is enabled.
  • Disabled by settings: for example, when MCP is disabled by settings, /mcp reports "MCP is disabled by settings"; when Skills is turned off by an admin policy, /skills reports it is disabled.
  • Varies by release: some commands are only available in certain releases or platforms.
  • Debug-only: some commands are visible only in development or debug mode and do not appear in normal use.
If you don't see a command, it usually means the corresponding capability is not enabled or has been turned off by settings. For the complete list of slash commands with aliases and scopes, see Slash Commands.

Built-in Subagents

A Subagent is a subordinate agent with its own context, dedicated prompt, and tool set. Qoder CLI CN comes with several built-in Subagents covering exploration, planning, memory, and Q&A. They are delegated automatically by the main session when needed, and can also be specified explicitly where supported.
SubagentPurposeCharacteristics
general-purposeGeneral research subagent for complex searches, multi-file analysis, call-chain tracing, and multi-step tasks.The default subagent; can use all tools (except tools exclusive to the main session).
ExploreFast read-only code exploration: find files by pattern, search keywords, understand existing implementations.Read-only; write and state-changing tools disabled; thoroughness can be specified per call (fast / medium / very thorough).
PlanRead-only solution design: lay out the implementation path, key files, dependency order, and architecture trade-offs.Read-only; outputs a step-by-step implementation plan listing the key files.
memory-managerManage cross-session memory, preferences, and facts; maintain personal and project instruction files.Available only when memory management is enabled; reads and writes memory files only.
statusline-setupConfigure a custom status line: read shell config, generate scripts, update settings.Interactive interface only.
qoder-guideAnswer questions about using and configuring Qoder CLI CN, Skills, Subagents, MCP, and Hooks.Answers primarily based on official docs; interactive interface only.
Built-in Subagents work without manual creation, and you can create your own Subagents with the qodercn agents command. For each built-in Subagent's tool set, model, and invocation, see Built-in Agents and Skills.

Built-in Skills

Skills package common workflows into reusable capabilities. Built-in Skills ship with Qoder CLI CN and can be invoked directly as slash commands (for example /simplify); some are also triggered automatically by the model in suitable scenarios.
SkillPurpose
/simplifyReview changed code for reusability, quality, and efficiency issues, launching multiple review perspectives in parallel and then fixing the findings.
/debugEnable debug logging for the current session, read the tail of the log, and help diagnose problems.
/security-scanRun a Qoder cloud security scan over the whole repository or a given path, supporting L2 lightweight and L3 deep review.
/questSmart workflow orchestrator that guides you through feature development step by step with dedicated subagents.
/mcp-configInteractively add, update, or remove MCP server configuration.
/loopRun a prompt or slash command repeatedly at a fixed interval.
/runLaunch and drive the project's actual application to observe changes at runtime.
/verifyVerify that code changes meet expectations by running the application and observing its behavior.
/rememberReview auto-memory entries, suggest promotions into instruction files, and identify stale, conflicting, and duplicate entries.
/batchApply bulk changes across many files by spawning parallel worker agents in isolated git worktrees.
/run-skill-generatorGenerate a project-level Skill for the project's run and debug steps.

How they trigger

Built-in Skills trigger in two ways:
  • Manual invocation: type the corresponding slash command in the interactive interface (such as /security-scan).
  • Automatic triggering: some Skills are used automatically when the model judges it appropriate; Skills marked as non-auto-triggering, such as /debug, /security-scan, and /batch, run only when you invoke them explicitly.

Conditional availability

A few built-in Skills depend on specific prerequisites:
  • /batch requires worktree isolation to be enabled (and the current directory to be a git repository).
  • /remember requires auto-memory to be enabled.
When the prerequisites are not met, these Skills do not appear in the available list. For the complete parameters and behavior of built-in Skills, see Built-in Agents and Skills; custom Skills can be managed with the qodercn skills command.

Workflows

Workflows orchestrate a series of steps into a repeatable process, ideal for solidifying multi-step tasks that recur within a team. Use /workflows to manage workflows, and /tasks and /kanban to track task progress. For detailed usage, see Workflows.

How to choose

  • Need a one-off management or auxiliary operation within the session → use a slash command.
  • Need to hand a class of tasks requiring its own context and standards to a focused executor → use a Subagent.
  • Need to reuse a mature workflow (review, debugging, security review, looped execution, etc.) → use Skills.
  • Need to solidify a multi-step task into a repeatable process → use a workflow.

Next steps