← ClaudeAtlas

scan-decklisted

Browse the deck — supportive triage view, filtered queues, kanban board, JSON dump, decision Q&A. Read-only by default. AUTO-INVOKE when user says "what's up?", "where do you need me?", "what's open", "show me the deck", "list cards", "kanban view", or "decisions to make". ALSO auto-invoke before Skill(create-card) to dedup against existing titles. Kanban first practice (make work visible).
zauberzeug/game-of-cards · ★ 2 · AI & Automation · score 68
Install: claude install-skill zauberzeug/game-of-cards
# Scan the deck Kanban's first practice (Anderson): **make work visible**. You cannot manage what you cannot see, and a swarm of /loop iterations cannot prioritize against state it has to reconstruct from chat. This skill renders the board for whoever's looking — human-supportive triage, filtered queues, kanban columns, JSON dumps, and a structured decision Q&A that closes the Andon-cord loop in one round. Read-only **except** in the explicit "decisions to make" mode, which calls `Skill(decide-card)` per parked card to lower gates. User argument: $ARGUMENTS ## Mode A — bare invocation / "what's up?" / "where do you need me?" The supportive default. Surface what's *blocking* progress (parked cards) before what's *queued* (open work), because the human's highest-leverage action is unblocking the line, not browsing it. ```bash goc triage ``` This emits parked cards (gate ≠ none) grouped by gate, oldest-first, with the `## Decision required` body section preview and an aged-days badge. Then show a one-line summary of the open queue: ```bash goc -v | head -10 echo "..." goc | wc -l ``` **Why `-v` by default**: the `-v` flag adds a per-card summary line that surfaces the qualitative context the `contribution` tier alone can't capture (pong-active vs pong-DORMANT, recent regression vs old doc-rot, blocking-other-work vs standalone). The agent picking from the queue makes better importance judgments with summaries visible — without them, "medium contribution" is opaque. End