Using agents in Visual Studio Code

Agents automate complete coding tasks that go beyond simple code suggestions and chat interactions. In Visual Studio Code, you can create agent sessions that run locally or in the cloud, interactively or in the background. Hand off tasks between different agent types to use their unique strengths. The unified Chat view gives you a central place to manage and monitor all your agent sessions, regardless of where they run.

This article provides an overview of the various agent types, how to create and manage agent sessions, delegate tasks between agents, and track their progress.

What are agents?

Agents perform complete coding tasks end-to-end. They understand your project, make changes across multiple files, run commands, and adapt based on the results.

For example, imagine you have a failing test. Instead of suggesting a fix, an agent can:

• Read the error message and identify the root cause across multiple files
• Update the relevant code
• Run the tests again to verify the fix works
• Commit the changes

You give an agent a high-level task, and it breaks the task down into steps, executes those steps with tools, and self-corrects when it hits errors.

You can run multiple agent sessions in parallel, each focused on a different task. When you create a new agent session, the previous session remains active, and you can switch between tasks via the agent sessions list.

Key concepts

The following concepts describe different aspects of working with agents in VS Code:

Types of agents

VS Code supports four main categories of agents, each designed for different use cases and levels of interaction:

Which agent should I use?

Use the following table to find the right agent type for your task:

Local agents

Local agents run directly within VS Code on your machine. You interact with local agents through chat to get immediate results to your prompts. Local agents have full access to your workspace, tools, and models. Use local agents for interactive tasks that require immediate feedback, such as brainstorming, planning, or exploratory work.

Local agent sessions use one of three built-in agents: Agent for complex coding tasks, Plan for creating structured implementation plans, and Ask for answering questions about your codebase. You can also create custom agents for specialized workflows.

Learn more about local agents in VS Code.

Background agents

Background agents, like Copilot CLI, are CLI-based agents that run non-interactively in the background on your local machine. They use Git worktrees to work isolated from your main workspace, preventing conflicts with your active work. Use background agents for well-defined tasks that have all necessary context, such as implementing a plan.

Learn more about background agents in VS Code.

Cloud agents

Cloud agents, like Copilot coding agent, run on remote infrastructure and integrate with GitHub repositories and pull requests for team collaboration and code reviews. Use cloud agents for well-defined tasks where you want to collaborate with team members through pull requests, or for tasks that can run without immediate feedback.

Learn more about cloud agents in VS Code.

Third-party agents

VS Code supports agents from third-party AI providers, like Anthropic and OpenAI. By using third-party agents, you can use the unique capabilities of these providers with your existing GitHub Copilot subscription, while benefiting from VS Code's unified session management and rich editor experience. Depending on the provider, third-party agents can run locally or in the cloud.

Learn more about third-party agents in VS Code.

Agent sessions list

The Chat view provides a unified view to manage all your agent sessions, regardless of where they run. By default, it shows your recent sessions, and gives information about their status, type, and file changes. Expand the list to see and filter all your agent sessions.

The list of sessions is scoped to your workspace. If you don't have a workspace open, the list shows all sessions across your workspaces. The sessions are grouped by time periods, such as Today or Last Week.

The Chat view operates in two modes: compact and side-by-side. You can manually switch between compact and side-by-side mode by using the toggle control in the top-right corner of the Chat view.

• Compact:

  In compact view, the list of sessions is embedded in the Chat view. When you select a session from the list, the Chat view switches to that session. Use the back button to return to the sessions list.

  

• Side-by-side

  In side-by-side view, the list of sessions is shown side-by-side with the Chat view. Select a session from the list to view its details in the Chat view.

  

  Tip

  When you make the Chat view wider, it automatically switches to side-by-side mode. Right-click on the sessions list and select Sessions Orientation to change this behavior (

  chat.viewSessions.orientation

  • Open in VS Code
  • Open in VS Code Insiders

  ). You can also use the toggle button.

Right-click a session in the list to see additional actions, such as different options to open the session details, archive the session, or agent-type specific actions like checking out a pull request (for cloud agent sessions).

To hide the session list from the Chat view, right-click in an empty chat and unselect Show Sessions (

Agent status indicator (Experimental)

The agent status indicator provides quick access to your agent sessions directly from the command center in the title bar. The indicator displays visual badges for unread messages and in-progress sessions, so you can stay informed about your AI agent activity without switching views.

The indicator shows:

• Unread sessions badge: Shows the count of chat sessions with new messages. Select the badge to filter the sessions list to show only unread sessions.
• In-progress sessions badge: Shows the count of sessions with running agents. Select the badge to filter the sessions list to show only in-progress sessions.
• Sparkle icon: Provides quick access to chat and session management options.

You can configure the indicator's behavior by using the

When a filter is active, the sessions list automatically expands to show all matching sessions. Select the badge again to clear the filter and return to the default view.

Create an agent session

You can create multiple agent sessions in parallel, each focused on a different task. When you create a new agent session, the previous session stays active, and you can switch between tasks through the agent sessions list.

When you create a new agent session, it starts with an empty context window. Each agent session is independent, so context from one session doesn't carry over to another.

You can create a new agent session from the Chat view or by using the corresponding commands in the Command Palette.

1. Open the Chat view and select the New Session dropdown (+).

   

2. Choose the agent type from the dropdown. Optionally, select a language model from the model picker.

   

3. Enter a prompt to assign a task to the agent. The agent starts working on the task.

   Generate a diagram that gives a high-level overview of the architecture of this project.
   

Hand off a session to another agent

You can hand off an existing task from one agent to another agent to take advantage of their unique strengths. For example, create a plan with a local agent, hand off to a background agent for proof of concepts, and then continue with a cloud agent to submit a pull request for team review.

To hand off a local agent session, select a different agent type from the session type dropdown in the chat input box. VS Code creates a new session, carrying over the full conversation history and context. The original session is archived after handoff.

In a background agent session, delegate to a cloud agent by entering the /delegate command in the chat input box. You can provide additional instructions after the /delegate command.

Assign a coding task to an agent

If you install the GitHub Pull Requests extension, you can assign an agent to implement TODO comments in your code.

On GitHub.com, or by using the GitHub Pull Requests extension, you can assign GitHub issues to Copilot coding agent by assigning the issue to copilot or by mentioning it in an issue comment or pull request to ask for a code review.

Review and apply file changes

When an agent session completes and makes code changes to your project, the session list shows the file change statistics for that session. To review the changes made by the agent, select the session from the list to open the session details.

Depending on the agent type, you have options to apply the changes made by the agent onto your local workspace, or to check out the branch from the agent session (for cloud agents).

Archive agent sessions

To keep the list of sessions organized, archive completed or inactive sessions. Archiving a session doesn't delete it but moves it out of the active sessions list. At any time, you can unarchive a session to restore it to the active sessions list.

To archive a session, hover over the session in the session list and select Archive. After you archive a session, it disappears from the list. Inversely, you can also unarchive a session in the same way.

To view your archived sessions, use the filter options in the sessions list and select the Archived filter.

Delete agent sessions

To permanently delete an agent session, right-click the session in the sessions list and select Delete. Deleting a session removes it permanently and can't be undone. For background agent sessions, deleting the session also removes any associated worktrees created for that session.

Related resources

• Agents tutorial: Hands-on tutorial for working with different agent types.

• Tools: Extend agents with built-in, MCP, and extension tools.

• Hooks: Execute custom commands at lifecycle events for automation and policy enforcement

• Custom agents: Create your own AI agents and extensions.