| name | resource-ops |
| description | Use when spawning, configuring, auditing, or retiring resource VMs that host infrastructure such as databases, services, test environments, or webhook-facing systems. |
Resource Operations
Use this skill when the task needs infrastructure rather than another disposable worker.
Placement Rule
When the task needs a new UI, service, database, webhook sink, or raw environment, first decide whether it belongs:
- inside Reef root as a control-plane module
- or on a separate VM as product/application infrastructure
Default to the separate VM path unless the feature is clearly part of Reef itself.
Do not use root as the default home for product services just because it is already running.
What a resource VM is
A resource_vm is passive infrastructure:
- database
- service
- build host
- test environment
- webhook sink
- other support system
It is not a token/cost usage entity.
Lifecycle default
resource_vm is protected-by-default.
Do not tear it down just because:
- the creating agent finished
- the parent subtree went historical
- the immediate task is done
Teardown should be:
- explicit user direction
- explicit owner policy
- or explicit root policy
Recommended flow
- spawn the resource VM
- SSH into it with
vers_vm_use
- configure and verify the service
- leave clear markers:
- ports
- processes
- file paths
- access instructions
- signal upward with enough detail for reuse or cleanup
What to report
When done setting up infrastructure, report:
- VM name
- VM ID
- what is running
- how to connect to it
- whether it should persist after this task
If you intentionally tear it down, say that it was explicit teardown, not incidental lifecycle cleanup.
