llms.txt

Deployment Runs

A deployment run tracks the execution of a deployment pipeline — from server provisioning through application startup. Every time you deploy or redeploy, a new run is created.

Viewing runs

Runs are displayed in the Runs tab on the right sidebar of the chat interface. Click on a run to view its detailed steps and logs.

The Runs tab in the right sidebar

Run steps

Each run consists of multiple steps executed sequentially.

Step statuses

StatusIndicatorDescription
PendingGrayStep hasn't started yet
RunningBlue, animatedStep is currently executing
CompletedGreenStep finished successfully
FailedRedStep encountered an error
SkippedGray, strikethroughStep was skipped

Real-time logs

Click on any step to view its real-time logs. Logs stream live during execution, showing the complete output of each command being executed on your server.

Run controls

Pause

Pause an active run to temporarily halt execution. The current step finishes, then the run waits. Useful when you need to investigate something before proceeding.

Resume

Resume a paused run. Execution continues from where it was paused.

Stop

Stop a run entirely. This cancels any remaining steps. Completed steps are not rolled back.

Run statuses

StatusDescription
StartedRun has been initiated
In ProgressSteps are actively executing
CompletedAll steps finished successfully
FailedA step encountered an unrecoverable error
PausedManually paused by user
StoppedManually stopped by user

Failed runs

When a run fails, the failed step is highlighted with error details and the AI will suggest fixes based on the error. You can:

  1. Fix the issue (e.g., update the port configuration)
  2. Ask the AI to modify the infrastructure plan
  3. Redeploy, which creates a new run

Redeployment

To redeploy your application after code changes:

  1. Push changes to your Git repository
  2. In the project chat, ask: "Redeploy the application"
  3. A new run is created with the latest code

Redeployments are faster than initial deployments because:

  • The server is already provisioned
  • Docker layers are cached
  • Only changed components need rebuilding

Run history

All past runs are preserved in the Runs tab. You can:

  • Review what happened in previous deployments
  • Compare successful and failed runs
  • Track deployment frequency and duration
  • Debug recurring issues by examining historical logs

Runs are ordered by most recent first, making it easy to find the latest deployment.