What should we prioritize next? The current biggest gaps

[thegdsks]

The roadmap tracks actual status, not aspiration. A lot of what was listed here as a gap has since shipped. Right now the biggest open gaps are:

• Real public ACME certificates, verified. TLS today defaults to an internal, self-signed issuer. A real Caddy ACME issuer, a settings toggle, and form validation are built and wired end to end, but issuance against a live domain has only been unit-tested at the config level, not spot-checked against a real domain.
• Multi-service apps, one remaining wizard gap. Backend fan-out and a real frontend (a services tab plus a deploy-spec form) are both shipped, and webhook-triggered auto-deploy fans out through the same path. The one remaining single-service-only surface is apps create --interactive, which always writes exactly one service; multi-service apps still need hand-written services: entries or the deploy-spec API/UI.
• Full team-based RBAC. Real multi-user accounts, per-user scoped abilities, curated named roles (admin/operator/viewer), and an IAM-style resource-scoped policy engine (Allow/Deny statements attachable to a user or token) all exist now, and there's a full audit log with CSV export. What's still missing is org-membership-based permissions, teams as their own first-class object.

Since the last pass, these have shipped: Docker Compose as a real deploy target, an install.sh with a systemd unit and a bounded post-install health check, and preview environments per pull request.

None of that is a secret backlog, it's what's actually next. What would be genuinely useful to hear: which of these is actually blocking you from trying Levelrail today, versus which ones you'd never have hit anyway. That's a better prioritization signal than we can generate internally.

Drop a comment on whichever gap affects you, or open a new one here if it's not on this list at all.