Three agents. One platform.

Every Hostking Pro Pod ships with a pre-installed agent stack. Pick the one that fits how you ship — or run all three side by side. They run isolated inside cgroups v2, so even a runaway agent cannot affect neighbouring tenants.

🤖Featured

Hostking Agent

Autonomous SysOps agent that audits Pod security, optimizes deployments, and runs nightly scans across every Castle tenant inside cgroups v2 boundaries.

›Inspects runtime metrics on each Pod and rebalances traffic
›Suggests tier upgrades when a Castle exceeds 80% PCU
›Can be triggered manually from the dashboard Console

🌿SysOps worker

Hermes Agent

Autonomous worker for deployments, cron jobs, and security. Hermes is your always-on pair-programmer that can ship to a Pod without human-in-the-loop approvals.

›Schedules cron jobs and webhook-triggered deploys
›Surfaces build failures with diff-level context
›Can run inside the same cgroup as the Castle it manages

🦞Personal assistant

Openclaw Agent

Your personal, always-on AI assistant that interfaces with terminal endpoints and deploys updates instantly. Ideal for solo founders and indie hackers.

›One-command deploys from your laptop
›Lives alongside your Castle tenants in the same Pod
›Bridges directly to the hostking CLI for power users

Built for AI-first teams

Hermes Agent and Openclaw Agent speak the same webhook protocol as the hostking CLI, so you can wire them into any pipeline.

Sandbox isolation

Every agent runs in a dedicated systemd unit with NoNewPrivileges=true, so even a compromised agent cannot escalate.

Launch Agent Console