workflows

To manage long-running application workflows, you can orchestrate persistent services by starting, stopping, restarting, and removing named processes. This approach ensures your services continue running and automatically restart after environment changes or dependency installations.

Monitoring status and port binding for persistent services is handled by configuring a named workflow with an optional port. The orchestration tool exposes APIs to retrieve real-time status and diagnose issues, ensuring your service remains properly bound and operational.

Configuring a start workflow involves specifying the execution command for your app and an optional port to monitor. This setup automates the lifecycle management of the process, allowing the service to persist, log activity, and restart automatically when needed.

Workflow orchestration requires a YAML frontmatter contract in the SKILL.md file, where name and description fields are mandatory. It also supports optional directories for scripts, references, and assets to structure your workflow configuration and dependencies.

Automatically restarting workflows after dependency installation or environment changes is a core feature. The orchestration mechanism monitors these updates and re-initializes the named processes to ensure the persistent services run with the latest configurations without manual intervention.

Long-running processes may fail to persist if the YAML frontmatter contract in SKILL.md is missing the required name and description fields. Ensure your workflow is properly configured with the correct start command and optional port binding to maintain service uptime.