The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
GitHub’s canvas pattern gives a recurring agent workflow a persistent, shared work surface for plans, status, drafts, decisions, and approvals. Rather than treating a long chat transcript as the only record, a canvas makes important state visible and editable by both a person and an agent. GitHub presents this as a way to make work easier to inspect and steer; its published examples do not independently measure time saved or productivity gains.
What GitHub means by a canvas
A canvas is a shared, interactive surface for a work artifact. Depending on the workflow, that artifact might be a plan, triage board, browser session, release checklist, dashboard, incident record, or spreadsheet. A person can edit the surface while an agent updates it, and the extension can expose actions the agent can call.
![The Microsoft Copilot Bible: [3 in 1] A Beginner's Guide to Harness AI in Office 365 to Automate Workflows, Generate Content, Analyze Data, and 10x Your Productivity](https://m.media-amazon.com/images/I/41+5Qv4pd2L._SL160_.jpg)
For example, GitHub describes an agentic kanban board with actions such as get_board, add_card, and move_card. The useful distinction is practical: related work is represented in a visible structure rather than existing only as messages to reconstruct later. GitHub’s description is a workflow-design argument, not evidence that spatial interfaces outperform chat in general.
Free tools Windows power users keep installed
One-click scans. No signup required.
Why chat alone can be hard to use for ongoing work
Chat is well suited to explaining intent, exploring ambiguity, and giving direction. But in a long conversation, execution details can become difficult to find: a plan may be followed by logs, corrections, validation results, and approvals, all embedded in the scroll. GitHub’s argument is that a canvas can give recurring work a stable home where someone can inspect the current artifact and status without replaying the whole conversation.
#1 Best Overall
That is GitHub’s product rationale, not a universal finding about every team or task. A short, one-off question may remain simpler in chat. A persistent surface is more relevant when work has stages, changes over time, or needs review and verification.
How to design a useful canvas
Ayan Gupta, a GitHub Developer Advocate, describes a four-part pattern for making agentic work visible and steerable:
- Define workflow states clearly. Name the meaningful stages so a person can tell where work stands.
- Surface decisions that matter. Make choices requiring judgment visible instead of burying them in conversation.
- Persist progress and drafts immediately. Keep the working artifact current as the agent and person make changes.
- Keep explicit human approval points. Identify where a person must review or authorize the next step.
The design principle is to represent the information people need to understand and decisions they need to make as explicit workflow state. Not every task needs this treatment: GitHub Docs positions canvases for workflows that need structure, iteration, and verification and where chat alone is insufficient.
What GitHub’s examples show
Java Modernization Studio
Gupta describes a canvas for Java modernization that represents assessment, planning, migration tasks, validation gates, and readiness to ship. The phases are visible and inspectable, letting reviewers focus on judgments while agents continue work between checkpoints. This is the author’s account of an example, not an independently evaluated case study.
Rank #3
Site Studio
Gupta also describes a canvas for managing personal-site content. It tracks section status, preserves drafts, supports iterative edits and review loops, and includes human approval points. The example illustrates how a persistent artifact can let someone steer or redirect work without rebuilding context at each iteration; it does not establish a measured improvement.
Together, the examples suggest a plausible fit for repeated, multi-stage work where tasks change hands or await review. Whether a canvas is worthwhile depends on the workflow, not on the novelty of the interface.
When a canvas is worth building
Before creating one, consider whether the workflow benefits from persistent, shared state enough to justify the setup and maintenance. These are practical evaluation questions based on the capabilities GitHub describes, not benchmark results:
- Does the workflow have states that need to remain visible across turns, sessions, or handoffs?
- Can both people and agents work with the same artifact rather than maintaining separate versions?
- Are decisions, validation points, and approvals easy to identify?
- Does the canvas reflect the real task and its artifacts, rather than adding a parallel status board?
- Is the workflow repeated or complex enough that a structured surface may offset the effort of building and maintaining it?
Gupta argues that a durable surface can reduce repeated prompting, context loss, back-and-forth, and rework on workflows that recur. The published article does not provide an independent evaluation, a measured time or cost reduction, or a quantified return on investment. Treat the proposed payoff as a potential benefit, not a guaranteed outcome.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to create a canvas in the GitHub Copilot app
- Start an agent session in the GitHub Copilot app.
- Enter
/create-canvasand describe the workflow, including the actions people and agents should be able to perform. GitHub’s example request is:Create an agentic kanban canvas with actions to create, assign, and move cards.
- Choose whether the extension should be available at project scope or personal scope. Project-scope extensions are stored under
.github/extensions; personal-scope extensions are stored under~/.copilot/extensions. - Describe the artifact and the workflow actions clearly. A canvas commonly includes a
package.json, an entry file such asextension.mjs, and optional JSON artifacts for persisted state.
GitHub Docs also gives examples such as a markdown canvas that combines meeting notes, prioritized issues and pull requests, and agent-session tracking. Issue-triage boards and document canvases are other documented use cases. The exact implementation and availability of product features can change; consult the current GitHub canvas extension documentation for the supported workflow.
What the reported build costs do—and do not—tell you
In an August 17, 2026 GitHub Blog article, Gupta reported spending about 2,000 AI credits on Site Studio and about 3,000 AI credits on the Java modernization canvas. These are the author’s reported build costs, not standardized benchmarks: the article does not establish a common pricing basis, a controlled comparison, or an independently measured return. It also does not quantify hours saved, percentage cost reductions, or throughput improvements.
For the author’s explanation and examples, see GitHub’s article on making agentic workflows visible and steerable.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
