| name | status |
| description | Report a Discord-formatted status dashboard of the fleet — master state, every worker's backend/model/uptime, and a summary footer. Trigger on the user typing "/status", "status?", "fleet status", "what's running", or anything semantically equivalent to "show me what's running right now." |
/status — fleet dashboard
When to use
The user wants a snapshot of fleet state — master plus every worker, their backends, and uptime. Trigger on any phrasing that maps to "show me what's running":
/status
status?
fleet status
what's running?
list workers
who's up?
What to do
Call the mcp__nanoclaw__list_workers MCP tool. The host runs the query synchronously and posts a formatted dashboard (master state + per-worker bullets + summary footer) to the master's channel as a chat message. The dashboard delivery itself counts as your reply — the host drops your trailing turn-text automatically. Just end the turn after the tool call; nothing else to do.
That's the entire skill.
What NOT to do
- Don't try to enumerate workers from
docker ps or your own context — list_workers is the only source of truth. It also includes archived workers and master state which docker ps doesn't.
- Don't paraphrase the dashboard. The format is intentional: the user reads it on a phone, scanning for icons + uptimes. Reformatting loses information.
- Don't run after every message. Only on explicit "show me status" intent.
v1 parity note
This is the v2 equivalent of v1 fleet's /status skill (FORK-SPEC §13.3), which ran tools/nc-status.sh and forwarded the output. The v2 form is simpler because the dashboard logic lives in list_workers itself rather than a separate shell script — the skill just routes the request.