Claude Agent
Claude Agent is a third-party coding agent by Anthropic bundled with AI Assistant. It understands your codebase, can plan and execute development tasks, and interact with your environment using tools, running commands, and analyzing their results to complete complex programming workflows.
Start using Claude Agent
Claude Agent is bundled with AI Assistant. You do not need to install it separately, it is available for use right away:
In the chat, click
and select Claude Agent from the list. 
Type your prompt and press Enter.
Accept the terms of service.
Once accepted, you can start using Claude Agent.
Use Claude Agent with an Anthropic API token
If you already have an Anthropic account, you can provide your API token in the settings and use Claude Agent with it.
Navigate to .
In the Third-party AI providers section, select Anthropic as Provider.
Enter your token in the API Key field and click Test Connection to verify that the connection is established.
Click OK to apply changes.
Switch authentication method
Claude Agent can be used either through a JetBrains AI subscription or with an Anthropic API key.
If you previously configured both authentication methods, the agent will continue using the one that was selected originally.
To switch the authentication method, you need to reset the current selection and then choose the desired one.
To do this:
In the chat, click the
Options button.Select the Reset ACP Authentication option from the list.
Start a new chat by clicking the
New Chat button or pressing Alt+Insert.Send a message to trigger the authentication prompt.
Select the authentication method you want to use.
The selected method will be saved and used for all future interactions until it is reset again.
Select operation mode
Claude Agent can operate in different modes, each providing a distinct level of autonomy and interaction.
To select an operation mode, click and select a mode from the list.
The following modes are available:
Default – standard behavior; prompts for potentially dangerous operations.
Accept Edits – automatically applies file edits.
Plan Mode – enables planning without executing actions.
Don't Ask – does not prompt for permissions and denies actions that are not pre-approved.
Bypass Permissions – bypasses all permission checks.
Select processing model
To select a model that Claude Agent uses to process your requests, click and select the model from the list.
Approve operations
By default, Claude Agent requests your permission to run suggested bash commands, perform file operations, or use external tools. In this case, you can either approve or skip the operation.
Always Allow – allows Claude Agent to execute this command automatically, without asking for permission.
Allow – allows Claude Agent to execute the command. You will be asked again if the command is requested later.
Reject – prevents Claude Agent from executing this command.
Rollback operations
If the changes introduced by Claude Agent do not suit you, you can roll them back. To do this:
Navigate to the pane listing the changed files.
Do one of the following:
To roll back changes in a specific file, hover over it and click
.
To roll back changes in all files, click Rollback.
You can also roll back to a checkpoint, restoring your project to an earlier point in the conversation to undo changes or retry with different instructions. To do this:
Navigate to the prompt you want to roll back to.
Hover over it and click
Rollback here.Confirm the operation by clicking Rollback.
This reverts all changes made by Claude Agent, as well as any manual changes you made after the selected point.
Enable use of external tools
You can enable Claude Agent to use tools provided by configured Model Context Protocol (MCP) servers, extending its capabilities to perform a wider range of tasks.
The available tools can be invoked automatically when the agent deems it necessary, or you can call them manually when writing a request.
To enable Claude Agent to use tools:
In the IDE settings (Ctrl+Alt+S), go to .
Enable the Pass custom MCP servers setting.
Click OK.
Use /commands
Claude Agent supports a subset of /commands for executing actions directly in the chat, along with bundled skills that extend its capabilities. For details on specific commands, refer to the official Anthropic documentation.
