eve-ng-node-operations

EVE-NG Node Operations

Use this skill for node lifecycle control.

Workflow

1. List or inspect nodes before acting.
2. Before starting or creating nodes, apply {baseDir}/references/node-guardrails.md.
3. For node creation choices, check {baseDir}/references/node-types.md.
4. Treat verified: false start/stop results as incomplete; confirm with status or console.

Available tools

• eve_list_nodes
• eve_get_node
• eve_create_node
• eve_delete_node
• eve_start_node
• eve_stop_node
• eve_start_lab
• eve_stop_lab
• eve_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.