Agent can execute terminal commands as part of its operations.

​
Command Types

Agent executes terminal commands for installing dependencies, running build processes, executing tests, starting development servers, and running scripts.

​
Features

​
Command Execution

Runs terminal commands with real-time output display. Supports multi-step command sequences.

​
Auto-run Mode

When enabled, commands execute without confirmation. Safety guardrails can be configured to restrict certain operations.

​
Output Monitoring

Agent monitors terminal output for error detection, process completion, and information extraction.

​
Configuration

Terminal settings available in Agent modes:
| Setting | Description | |:--------|:------------| | Enable Terminal | Allow Agent to execute terminal commands | | Auto-run | Execute commands without manual confirmation | | Auto-fix Errors | Automatically attempt to resolve command failures | | Allow lists | Configure which commands are permitted to run automatically |
Terminal capabilities can be configured separately for each Agent mode.