AIToday
Large Language ModelsAI Coding AssistantsGitHub Copilot BlogPublished: Aug 7, 2026, 06:00 JST4 min read

GitHub Copilot app adds slash commands for faster workflows

GitHub Copilot app adds slash commands for faster workflows

Key takeaway

  • GitHub Copilot's desktop app now includes slash commands that function as workflow shortcuts in the chat composer—typing / triggers an autocomplete menu showing available commands for your current context.

  • Key commands include /plan (for breaking down tasks), /spar (for pressure-testing ideas), /autopilot (for multi-step implementation), /rubber-duck (for independent review by a different model), /create-canvas (for interactive visualizations), and /orchestrate (for coordinating changes across repositories).

  • Unlike CLI slash commands, which focus on terminal operations, the app commands are tailored to the visual, multi-session workflow the desktop interface provides.

3 Key Points

  1. What happened

    GitHub has introduced slash commands in the GitHub Copilot app—text shortcuts that appear when you type / in the chat composer. Commands include /plan (to break down tasks before coding), /spar (to challenge your approach), /autopilot (to automate implementation steps), /rubber-duck (to get a second opinion from a different model), /create-canvas (to turn conversations into interactive interfaces), and /orchestrate (to coordinate work across multiple repositories).

  2. Why it matters

    Slash commands let developers move faster and stay focused by replacing manual navigation with quick text shortcuts. The app-specific commands differ from the CLI version because the desktop app provides a visual interface for managing project context automatically—so commands focus on workflows (like navigating sessions and managing projects) rather than terminal-level tasks like setting directories.

  3. What to watch

    The /plan and /autopilot commands appear to form a core workflow—plan first, then automate implementation. The /rubber-duck command uses a different model to independently review work, which may help surface blind spots. GitHub's docs contain additional examples and details on creating canvases and using orchestration for multi-repository changes.

In Depth

Read the full story

GitHub has introduced slash commands in the GitHub Copilot app, a set of text shortcuts designed to streamline common developer workflows. Users access these commands by typing / into the chat composer, which triggers an autocomplete menu showing the available commands for their current context.

The key difference between app and CLI slash commands stems from their respective environments. CLI slash commands are designed around a terminal-first workflow, handling tasks like adding directories, setting the working directory, and managing terminal access. The app, by contrast, provides a visual interface that automatically manages project context, so its slash commands focus on workflows instead. Some commands like /clear and /model work in both places, but app-specific commands are tailored to the multi-session, desktop-based workflow.

The slash command set supports a progression through development activities. /plan helps developers break down a task before writing code, think through their approach, identify potential challenges, and decide what happens next—it also switches the session into Plan mode. Examples include planning a new feature by identifying files and dependencies, mapping out a complex refactor before touching the code, or triaging a bug to explore possible causes. /spar acts as a pressure-testing tool, using Copilot to question assumptions and point out potential risks or tradeoffs, such as challenging a caching strategy, debating REST versus GraphQL, or identifying edge cases in a migration plan. /autopilot takes the wheel once a plan exists, handling multi-step implementation tasks and allowing developers to give Copilot a goal rather than managing each individual step—examples include implementing a new feature or updating dependencies. /rubber-duck provides an independent review using a different model, surfacing blind spots and catching issues the primary model may have missed; it is positioned for reviewing complex plans, large refactors, or migration strategies before moving forward.

/create-canvas transforms conversations into interactive experiences by creating visualizations, dashboards, or custom workflows that developers can interact with directly, rather than working through information in a chat window. /orchestrate coordinates work across multiple repositories when a feature touches multiple codebases or when several related tasks must move forward in parallel. The article notes that users do not need to memorize commands—the autocomplete menu provides guidance, and additional examples and documentation are available in GitHub's blog and official docs.

Context & Analysis

The GitHub Copilot app's slash command system reflects a shift in how AI coding assistants integrate into developer workflows. The article makes clear that the app's commands differ fundamentally from the CLI version because the environments are different: the CLI commands manage terminal-level tasks (directories, working directory, terminal access), while the app provides a visual interface that handles project context automatically. This design choice suggests GitHub is optimizing slash commands for the strengths of each interface—terminal-first for the CLI, workflow-first for the desktop app.

The command set reveals a deliberate progression through common development activities. Developers begin with /plan to break down a task, then use /spar to stress-test the approach before committing, then move to /autopilot to automate implementation. The /rubber-duck command is positioned as a quality gate—a way to catch blind spots before pushing code. Finally, /orchestrate handles the multi-repository coordination that larger features often require. This sequence suggests the commands are meant to support a complete feature lifecycle, not just isolated tasks.

The mention that /clear and /model commands work in both CLI and app versions indicates GitHub is building a consistent Copilot vocabulary across platforms where it makes sense, while allowing context-specific commands to flourish where they add the most value.

FAQ

How do I access slash commands in the GitHub Copilot app?
Type / directly into the GitHub Copilot app's chat composer, and an autocomplete menu will appear showing the commands available in your current context.
How do /plan and /autopilot work together?
You use /plan to break down a task before you start writing code, think through your approach, and identify potential challenges. Once you have a plan, /autopilot helps you turn that idea into working code by managing the implementation steps and iterating as needed.
What does /rubber-duck do differently from the main model?
It uses a different model to independently review your work, helping surface blind spots, question assumptions, and catch issues your primary model may have missed—providing a second opinion before you move forward.
GitHub Copilot BlogRead Original Article

Get the latest Large Language Models news every morning

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytime

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleAgentsview tool tracks costs across AI coding agents locally

The AI news that matters, in one minute each morning.

Sign up free