App hosting and deployment on Resilis

Your editor or external coding agent creates the application source. Resilis builds and runs the selected source, applies the configuration you review, and reports deployment status. Connecting an agent gives it access to supported cloud tools through the Model Context Protocol, or MCP. It does not give the agent unrestricted permission to change production.

You choose the organization, project, environment, and compute target. Keep development and production settings distinct. Review the source revision, required variables, data dependencies, and target before starting a deployment. A successful build is one checkpoint; you must also check that the running application serves the expected result.

Use the dashboard when your source is in GitHub. Connect a repository, select its branch and application root, and review the build settings. The existing Node.js guides show a recorded GitHub deployment with Railpack and a separate Dockerfile deployment. Use those examples to check the workflow before adapting your own application.

Use a connected agent when the source is on your computer. This route needs Git and an agent with local file and command access. The agent prepares a reviewed commit and transfers source to a private Resilis-hosted Git repository. GitHub is unnecessary for this route, but source still leaves your computer. Keep secrets out of uploaded files.

• Deploy your first app →
• Deploy a Node.js app with Railpack →
• Deploy a Dockerfile →

Managed database workflows support PostgreSQL and MySQL. Create the database on a compatible target and bind its connection to the app as a sensitive environment variable. Database persistence is separate from the application process: an in-memory task board loses added tasks when the process restarts, while the PostgreSQL example checks that a task survives an app rebuild.

Managed object storage provides buckets and objects through an S3-compatible backend. You can also connect an existing Amazon S3, Cloudflare R2, or Backblaze B2 bucket. That connection registers an accessible bucket; it does not create the bucket or migrate existing objects. Provider credentials and permissions still apply.

• Managed databases, recovery, and object storage →
• Node.js with PostgreSQL example →

Run on Resilis compute, or connect an existing Linux VPS or Kubernetes cluster through the Resilis connector. Connected compute remains customer infrastructure. Supported operations depend on target health and the capabilities it advertises. Connecting a target does not automatically adopt or migrate applications already running there.

Custom domains are not available on connected compute or S3-backed static sites. S3-backed static sites also do not provide container runtime logs. Autoscaling is not part of the current app-creation path. Check the selected resource and target before relying on a feature described for managed application hosting.

• Deployment settings and connected compute →

Use deployment status, build logs, and supported runtime logs and metrics to investigate a release. Open the serving URL and perform a real application action. If a release fails, inspect its status and logs before changing configuration or retrying. Restart, redeploy, and rollback are available only where the resource and target support them.

App rollback changes the application release. It does not restore database data or reverse schema migrations. Review database protection independently: Hobby includes snapshots, while Pro and above include point-in-time recovery entitlement for supported, configured database paths. Available archives and restore points determine what can be recovered.

Hobby costs $10 per month and Pro costs $50 per month. The plan fee becomes runtime credit for compute, persistent disk, database fees, egress, and object storage. Resource usage and build overage can affect the bill. Read the rate table and scoped example before choosing capacity; bucket limits are not an allowance of free object-storage capacity.

• Agent access and authorization →
• Usage rates and worked example →
• Compare plans →