EVE-NG Node Operations
Use this skill for node lifecycle control.
Workflow
- List or inspect nodes before acting.
- Before starting or creating nodes, apply
{baseDir}/references/node-guardrails.md. - For node creation choices, check
{baseDir}/references/node-types.md. - Treat
verified: falsestart/stop results as incomplete; confirm with status or console.
Available tools
eve_list_nodeseve_get_nodeeve_create_nodeeve_delete_nodeeve_start_nodeeve_stop_nodeeve_start_labeve_stop_labeve_wipe_node
Node-ID and runtime-status guardrails
- Preserve existing EVE node IDs; do not renumber or replace node IDs as a workaround unless the user explicitly asks.
- On local EVE, treat node runtime as lab-UUID scoped (
/opt/unetlab/tmp/0/<lab_uuid>/<node_id>). EVE API status can leak across labs that reuse node IDs, so use MCP-corrected status or verify local runtime before concluding a node is up. - If a same-ID node in another lab appears UP, fix the runtime/status scoping problem; do not rewrite the lab's node IDs.