Background sessions and agent view

Agent view is a screen in Claude Code that lists every session on your machine as one row each, showing what is running, what has finished, and what is waiting on you.

Last updated: 29 Sep, 2026 · Claude Code

When there are four things running, the problem stops being how to start them and becomes how to see them.

The video starts from the problem. Each new task means another terminal and another claude: one to update the README, one to write unit tests. Agent view dispatches and manages many sessions from one screen instead. claude agents opens it, with the sessions grouped as Needs input, Working and Completed.

The video types a task into the input at the bottom, "please update the readme file with the topic agent view in claude code", then a second, "please check the unit test cases created in test_example.py". Both run in parallel. The test task moves to Needs input when it wants permission to read a file; the video opens it, approves, and goes back to the list with /bg.

A subagent can be dispatched the same way. The video types "Run the code improvement advisor agent" from the subagents lesson; it works in the background beside the other tasks, and when it finishes its row moves to Completed. Opened, it reports that it reviewed all three Python files without editing them, lists the top findings, and asks which fixes to apply.

claude agents

Agent view is one row per session. It shows what is running, what has finished, and, most usefully, which sessions are stuck waiting for a permission answer. You attach to a row to drop into that conversation, and leave it again.

What it is for

Dispatch a bug fix, a review, and an investigation into a flaky test. Go and do something else. Come back to three rows: two finished and one asking a question. The screen is built for this kind of work, and it helps most when the tasks are independent.

• Independent jobs, not steps of one job.
• Work you are willing to check afterwards rather than watch.
• Tasks small enough to describe in a sentence, because you are not there to steer.

Which raises the obvious question: how do you trust work you did not watch? Everything in part 4 is the answer. Permission rules, a sandbox, and hooks are what make a background session something other than a gamble.

Sessions that talk

Two other features sit near this one. Cross-session messaging lets one session pass findings to another. Agent teams put a lead session in charge of teammates that talk to each other, which is the next lesson: Agent teams.

Related

• Previous: Worktrees
• Next: Agent teams
• Reference: Claude Code docs