60-second Chump fleet operator briefing — 24h ship count and rate, pillar mix from shipped PR titles, open PR stalls (BLOCKED > 4h), auto-fixed CI events, manual rescues, and a suggested next operator action. Use when starting a session and wanting a quick "what's the fleet doing" read, OR when an alert has fired and you want context before diagnosing. Thin wrapper over harness-neutral CLI at `scripts/dispatch/fleet-brief.sh` (INFRA-721); same call works from any harness. Per `.claude/README.md` pattern — capability lives in the script, this skill just exposes it as a slash command.
Install
npx skillscat add repairman29/chump/fleet-brief Install via the SkillsCat registry.
/fleet-brief — Chump Fleet 60-Second Briefing
Canonical surface: `scripts/dispatch/fleet-brief.sh` (INFRA-721). Any harness invokes the same script; this skill is the Claude Code adapter.
Routing
Arguments passed: $ARGUMENTS (typically empty — the brief takes no required args).
scripts/dispatch/fleet-brief.sh $ARGUMENTSSurface stdout to the user verbatim. The script outputs ≤ 30 lines of plain text designed for scannability — don't re-paraphrase.
When the user asks "what's the fleet doing?"
This is the right tool. The brief covers:
- 24h ship count + rate trend
- Pillar mix from shipped PR titles (RESILIENT / EFFECTIVE / CREDIBLE / ZERO-WASTE / MISSION)
- Open PR stalls (BLOCKED > 4h)
- Auto-fixed CI events (lint / flake reruns) — count of saved operator interventions
- Manual rescue events
- Suggested next operator action
When NOT to use this
- For pure health pass/fail (use
/fleet-doctor— exit code is the answer) - For "is feature X shipped?" lookups (use
/verify-existence) - For alert response (use
/operator-recall --check-onlyfirst)
Related capability
The SessionStart hook (FLEET-019) already invokes fleet-brief.sh at session start, so the briefing shows up in your session prompt automatically. This skill exists for on-demand re-runs — when you want a fresh read mid-session.