Guide intermediate

Cursor 3.2: /multitask and Canvases Turn the IDE Into an Agent Execution Platform

Published April 29, 2026 · Updated May 1, 2026 · by Jonathan Hildebrandt

Cursor 3.2 ships /multitask for parallel async subagents and interactive Canvases that replace chat walls with live dashboards. Here's what changed and how to use it.

Table of Contents
Pondero, operated by Hildebrandt AI LLC, earns a commission from some links on this page. This does not influence our editorial decisions. Read our affiliate disclosure

Cursor 3.2: /multitask and Canvases Turn the IDE Into an Agent Execution Platform

In brief: Cursor 3.2 (released April 24, 2026) adds /multitask for spawning async parallel subagents and interactive Canvases that let agents build live dashboards instead of dumping text into chat. If you’re doing anything beyond single-file edits, this changes how you work.

What Changed

/multitask: Parallel Async Subagents

Before 3.2, if you queued multiple requests in the Agents Window, Cursor processed them serially. One task blocked the next. With /multitask, Cursor spawns a fleet of async subagents that run simultaneously. Larger tasks get automatically broken into chunks and distributed across subagents. The tiled layout in the Agents Window lets you watch outputs side-by-side without switching tabs.

Practically: if you ask Cursor to refactor three modules and write tests for each, it no longer finishes module one before starting module two. All three run in parallel.

Canvases: Interactive Artifacts in the Side Panel

Canvases are durable, interactive outputs that live in the Agents Window side panel alongside the terminal, browser, and source control. Instead of getting a wall of markdown in chat, agents can now build:

  • Tables and charts with real data
  • Diagrams and flowcharts
  • Diff views and to-do lists
  • Custom interactive interfaces with logic baked in

Canvases persist across the session. You can explore and interact with them (sort a table, drill into a chart) rather than re-prompting for updates.

Expanded Worktrees and Multi-Root Workspaces

Cursor 3.2 also expands worktrees in the Agents Window for isolated background work across branches. A single agent session can now target a multi-root workspace spanning multiple repos, which means cross-repository changes no longer require juggling separate windows.

Why It Matters

These features push Cursor past “AI-assisted coding” into something closer to an agent execution runtime. The practical shift:

  • Parallelism eliminates the biggest bottleneck in agentic workflows: waiting. Teams running multi-module refactors, test generation, or documentation passes will see wall-clock time collapse.
  • Canvases fix a real UX problem: markdown chat output is hard to scan and easy to miss. A live dashboard for a migration plan or dependency graph is immediately more useful than 400 lines of text.
  • Multi-root workspaces matter most for monorepo teams or anyone who regularly spans services. Asking an agent to make a coordinated change across a frontend and backend repo, and see it happen, is now a single session.

For operations teams and developers managing large codebases, this release meaningfully lowers the overhead of running agentic tasks at scale.

How to Use It

Getting started with /multitask:

  1. Open the Agents Window (Cmd+Shift+A on Mac)
  2. Type /multitask followed by your compound request, e.g., /multitask refactor auth module, add unit tests, and update the README
  3. Cursor breaks the request into subtasks and spawns subagents. Watch them run in the tiled pane layout.

Working with Canvases:

  1. In the Agents Window, ask for output that benefits from structure: dashboards, comparisons, architecture diagrams
  2. The agent will automatically create a Canvas in the side panel rather than dumping text into chat
  3. Interact directly: click rows, adjust filters, or ask the agent to update the Canvas in-place

Multi-root workspaces:

  1. Open multiple repo folders in a single Cursor window (File > Add Folder to Workspace)
  2. In the Agents Window, your agent session has access to all roots, so you can reference files across repos by path

Pricing note: /multitask and Canvases are available on the Pro plan ($20/mo) and above. The Business plan ($40/user/mo) adds self-hosted cloud agent support announced in the same release cycle.

If you’re evaluating Cursor against its closest competitor, see our Cursor vs Copilot comparison. For a broader look at the field, the Best AI Coding Tools guide covers the full landscape.

You can also read our Cursor full review for a breakdown of strengths, weaknesses, and who it’s actually right for.

This post is part of Pondero’s daily coverage of AI tool updates. See all coding guides →