What’s new in DeployPro. Upgrade by running the installer again.
Claude explains failed deploys
A failed deployment now has an Explain this failure button. Claude reads the end of the build log and the project’s settings, then says what broke, why, and what to change, with how confident it is. Every variable’s value is replaced by its name before anything is sent. An admin turns it on once under System → AI assistant, with the team’s own Anthropic key.
An Overview of the whole server
A new first page: memory and CPU of the server, memory claimed by every project’s limits, the projects using the most right now, what needs attention, recent deploys, disks, backups and the latest changes.
Also
A health check per project: path, timeout, and whether an error page counts as healthy.
Builds wait for free memory instead of risking running apps.
Notifications per project to Slack, Discord and Microsoft Teams: deploys succeeded or failed, outages and recoveries.
Run a published image (n8n, MinIO, Uptime Kuma…) as a project, no repository needed.
A redesigned sign-in page with both ways in, and a new DeployPro mark.
Teams, two-factor sign-in and an audit log
Accounts with roles: viewer, developer, admin and owner. Invitation links from the Team page.
Two-factor sign-in with any authenticator app, and recovery codes.
Personal API tokens with read, deploy or full scope and an expiry.
An audit log of every change and failed sign-in: who, what, when, from where.
Logs, graphs and a livelier dashboard
Live runtime logs for the website, every worker and Redis, with filter, errors only, pause and download.
CPU, memory and network graphs per project over an hour, a day or a week.
Edit, reveal, paste and download environment variables.
Statuses update without reloading; confirmations open as dialogs; success shows as a toast.
A ✓ or ✗ on each GitHub commit and the preview’s address on its pull request.
Background services get a fixed internal address, <project>.internal, for shared services such as ClamAV.
Built-in Redis and background services
A Redis per project for job queues and caches, kept running and backed up.
Projects that run with no website: queue consumers built from their own Dockerfile.