--- title: GitHub Copilot desktop app and canvas feature overview description: GitHub shipped a standalone Copilot desktop app for Windows and Mac that goes beyond code completion, handling issue tracking, PR review, isolated work trees,... canonical: https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: GitHub Copilot desktop app and canvas feature overview | daily.dev og:description: GitHub shipped a standalone Copilot desktop app for Windows and Mac that goes beyond code completion, handling issue tracking, PR review, isolated work trees,... og:url: https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r og:image: https://api.daily.dev/og/posts/mJ0gdeU2R.png og:image:alt: GitHub Copilot desktop app and canvas feature overview og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # GitHub Copilot desktop app and canvas feature overview **[Collections](https://daily.dev/sources/collections)** · 4 min read · 4 upvotes · 0 comments ## Summary GitHub shipped a standalone Copilot desktop app for Windows and Mac that goes beyond code completion, handling issue tracking, PR review, isolated work trees, and local or cloud automation tasks. A new `/create-canvas` slash command turns agent conversations into shareable, interactive micro-apps (dashboards, checklists, internal tools) without requiring separate hosting. Canvases can be scoped to a project via `.github/extensions` or kept personal in `~/.copilot/extensions`, and consist of a package.json, an entry file, and an optional artifacts directory. Existing canvases can also be installed via plugins like Azure DevOps or reused from a repo's Installed list. ## Content ## What is "Continuous AI"? GitHub Next coined the term to describe applying the same philosophy behind CI/CD to AI-driven work: event-triggered, repetitive, auditable tasks that run without someone manually kicking them off. Think triage, documentation updates, fault analysis, daily summaries. The GitHub Copilot app's Automations feature is the first practical place to try this. ## The GitHub Copilot desktop app The app (available at github.com/github/app for Windows and Mac) is where most of this started. You define an automation with a prompt, a trigger (schedule, issue event, or PR event), a choice of local or cloud execution, and which tools Copilot can use. It supports coding tasks, issue tracking, PR review, and isolated work trees. Adoption has apparently spread beyond developers. GitHub's legal team reportedly uses it, which says something about how low the barrier to entry is. ### A simple example: morning summary One walkthrough example is a "where was I?" automation that runs each morning and summarizes what happened in a repo overnight. You write the prompt, set a daily schedule, and let it run. The app handles the rest. On prompt injection: the recommended mitigation is restricting write access on triggering events, so an attacker can't craft an issue that causes the automation to do something harmful. ## VS Code gets automations too The same concept landed in VS Code's Agents window as a preview feature, and has since shipped as a default. To set one up, you need `chat.automations.enabled` turned on (now on by default), then you define: - A prompt - A workspace or worktree target - Which agent to use - Session mode - A schedule: Manual, Hourly, Daily, or Weekly A few operational caveats worth knowing: the host machine has to stay running, missed runs aren't guaranteed to catch up, and only one session runs at a time per automation. Compared to the Copilot app, VS Code currently only supports time-based triggers. No CRON expressions, no issue or PR triggers yet. ## Sharing automations Once you've built something useful, you can share it. VS Code supports three approaches: **Agent plugins** — bundle automations as templates inside a plugin, placed in an `automations/` folder as `.automation.md` files with YAML frontmatter and a Markdown prompt. Good for team-maintained templates that update centrally. **Export/import** — export a single `.automation.md` file and send it to someone. Good for one-off sharing. **What gets shared vs. what stays local** — only the portable parts travel: name, prompt, schedule, format version, and identifier. Workspace, provider, model, permissions, enabled state, and run history are always set locally by whoever receives it. Unsupported schedule values are rejected outright rather than silently changed. ## A practical example: weekly skill audit One developer ported the `skill-stocktake` slash command from the ECC repository (originally built for Claude Code) to run as a scheduled Copilot automation. It audits installed AI agent skills weekly, checking for overlap, staleness, and outdated technology references, then produces verdicts: Keep, Improve, Update, Retire, or Merge. The Copilot port differs from the original in a few ways: it uses plain instructions instead of scripts or subagents, infers usage from repository evidence rather than usage counts, and never modifies files without explicit confirmation. ## Canvases: sharing the output If an automation produces something people need to interact with, the `/create-canvas` slash command in the Copilot app turns an agent conversation into a shared interactive interface — a dashboard, checklist, or similar. No hosting or deployment required. Canvases are bidirectional: the agent updates them while working, and users can click and interact directly. They can be scoped to a project (committed to `.github/extensions`, shared with the whole team) or kept local per user (`~/.copilot/extensions`). Under the hood, an extension is a package with a `package.json`, an entry file (`extension.mjs`), and an optional artifacts directory. Existing canvases can also be installed via plugins or opened from a repo's Installed list without recreating them. ## Where things stand The VS Code implementation is newer and more limited than the Copilot app version, but it's moving fast. Time-based triggers work now; event-based triggers are presumably coming. The sharing infrastructure is already solid enough to distribute automations across a team without much friction. The interesting thing isn't any single feature here. It's that the pattern — write a prompt, attach a trigger, let it run on a schedule — is becoming a first-class primitive in the tools developers already use. ## Questions this post answers ### What is the GitHub Copilot desktop app and what can it do beyond code completion? The GitHub Copilot desktop app is a standalone application for Windows and Mac available at github.com/github/app that extends beyond coding assistance to handle issue tracking, pull request review, and isolated work trees. It also runs local and cloud automation tasks, positioning it as a broader workflow tool rather than just an in-editor assistant. _Developers evaluating new AI coding tools can track releases like this one on daily.dev._ ### How does the /create-canvas command work in GitHub Copilot? The /create-canvas slash command turns an agent conversation into a shareable, interactive interface such as a dashboard or checklist, without needing separate hosting. Canvases are bidirectional, meaning the agent updates the canvas as it works while users can click and interact with it directly, and they can be scoped to a project via .github/extensions or kept personal in ~/.copilot/extensions. _Teams adopting agent-driven internal tools can follow how features like Copilot canvases evolve on daily.dev._ ### What files make up a GitHub Copilot canvas extension? A canvas extension consists of a package.json file, an entry file named extension.mjs, and an optional artifacts directory. Extensions can be installed through plugins such as Azure DevOps or opened directly from a repository's Installed list, avoiding the need to rebuild the same tool repeatedly. _Anyone building or reusing Copilot extensions can keep tabs on tooling changes via daily.dev._ ## Similar posts on daily.dev - [GitHub Copilot tutorial: How to build, test, review, and ship code faster \(with real prompts\)](https://daily.dev/posts/github-copilot-tutorial-how-to-build-test-review-and-ship-code-faster-with-real-prompts--lyeseh4bm) · GitHub Blog · 2 upvotes · 0 comments - [Getting started with GitHub Copilot in Visual Studio or VS Code](https://daily.dev/posts/getting-started-with-github-copilot-in-visual-studio-or-vs-code-3or4rwi6m) · InfoWorld · 0 upvotes · 0 comments - [Automate your full development process with handoffs](https://daily.dev/posts/automate-your-full-development-process-with-handoffs-1xchwgu3j) · The Art of Simplicity · 1 upvotes · 0 comments --- Tags: [#devtools](https://daily.dev/tags/devtools), [#github](https://daily.dev/tags/github) [View this post on daily.dev](https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"GitHub Copilot desktop app and canvas feature overview","url":"https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r"},"datePublished":"2026-09-16T05:56:52.803Z","dateModified":"2026-09-22T05:08:02.027Z","description":"GitHub shipped a standalone Copilot desktop app for Windows and Mac that goes beyond code completion, handling issue tracking, PR review, isolated work trees,...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f37b2a4ea2f0bc8a325787c91ef71903?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/f37b2a4ea2f0bc8a325787c91ef71903?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":4},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"devtools,github","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"GitHub Copilot desktop app and canvas feature overview"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/github-copilot-desktop-app-and-canvas-feature-overview-mj0gdeu2r#faq","mainEntity":[{"@type":"Question","name":"What is the GitHub Copilot desktop app and what can it do beyond code completion?","acceptedAnswer":{"@type":"Answer","text":"The GitHub Copilot desktop app is a standalone application for Windows and Mac available at github.com/github/app that extends beyond coding assistance to handle issue tracking, pull request review, and isolated work trees. It also runs local and cloud automation tasks, positioning it as a broader workflow tool rather than just an in-editor assistant. Developers evaluating new AI coding tools can track releases like this one on daily.dev."}},{"@type":"Question","name":"How does the /create-canvas command work in GitHub Copilot?","acceptedAnswer":{"@type":"Answer","text":"The /create-canvas slash command turns an agent conversation into a shareable, interactive interface such as a dashboard or checklist, without needing separate hosting. Canvases are bidirectional, meaning the agent updates the canvas as it works while users can click and interact with it directly, and they can be scoped to a project via .github/extensions or kept personal in ~/.copilot/extensions. Teams adopting agent-driven internal tools can follow how features like Copilot canvases evolve on daily.dev."}},{"@type":"Question","name":"What files make up a GitHub Copilot canvas extension?","acceptedAnswer":{"@type":"Answer","text":"A canvas extension consists of a package.json file, an entry file named extension.mjs, and an optional artifacts directory. Extensions can be installed through plugins such as Azure DevOps or opened directly from a repository's Installed list, avoiding the need to rebuild the same tool repeatedly. Anyone building or reusing Copilot extensions can keep tabs on tooling changes via daily.dev."}}]} ```