Your agent control plane
Your VPS. Every agent.
See what is running, deploy Hermes or OpenClaw in one click, and give every agent a saved persona and project memory.
01 · Connect
See the box you own.
ACRE reads the live health, capacity, uptime, and resident state of your runtime. No mystery infrastructure.
Live runtimeOne honest status surface
02 · Deploy
Pick an agent. Press deploy.
Choose Hermes, OpenClaw, nano-claw, or MCP. ACRE creates the resident, attaches project memory, and routes it through your runtime.
One clickRuntime adapter included
03 · Shape
Every agent keeps a self.
Name it, write its persona, and set standing instructions once. That profile returns with the agent in every session.
Saved profilePersona · role · instructions
04 · Remember
Memory stays with the project.
Every resident reads the same confined project brain. Useful findings survive sessions without bleeding into another acre.
Durable memoryShared inside one boundary
05 · Manage
Start. Stop. Restart.
Inspect each resident, message it, or change its runtime state from one place—while the control plane stays on your infrastructure.
Your groundRuns on your VPS
The result
Your agent stack becomes legible.
One runtime becomes a living acre. Every agent has a visible job, durable context, and a place you can return to.
Walk the acre
A control plane you can understand.
ACRE sits above your agents and your VPS. It turns runtime health, deployment, persona, and memory into one clear surface without hiding the machine underneath.
Connect the runtime
Run ACRE on a Linux VPS. The app reports real host health and resident state through the same-origin API.
Deploy a resident
Choose a runtime adapter, give the agent a name and persona, then deploy it directly into the active acre.
Manage the work
Chat, inspect, start, stop, and restart residents while their standing instructions and project memory persist.