Multitask

What it means

The product queues or runs concurrent agent sessions, each with its own context, tools, and progress, while the user continues working elsewhere.

Why designers should care

Multitask UX needs a task list, status per job, conflict warnings when two agents touch the same file, and clear completion signals—not silent background churn.

Example

Cursor Multitask runs lint fixes on branch A while the user chats on branch B; a sidebar shows both jobs with Stop, Open, and Merge when each completes.

Common mistakes

• • Parallel agents editing the same file with last-write-wins and no merge UI.
• • No token or cost visibility across concurrent runs.
• • Background jobs that finish with no notification or summary receipt.

Related terms

• Background Agent: The agent session decouples from the foreground UI: work continues with its own logs, tools, and time limits until done, failed, or cancelled.
• Orchestration: An orchestrator decides what runs next: call search, pass results to the LLM, invoke a tool, branch on validation, or pause for approval.
• Worktree: Each worktree has its own branch and file tree; commits in one do not block edits in another until merge, reducing conflict during parallel agent runs.
• Fork: The product copies context up to a message or step into a new thread/session, letting users pursue alternate prompts, edits, or agent plans in parallel.
• Token Burn Rate: Tokens in plus tokens out, multiplied by call frequency and retries, equals burn rate. Long context, verbose prompts, and chained inference multiply it quickly.