agentleFS
zauberzeugzauberzeug/game-of-cards/.claude/skills/standup/SKILL.md

standup

Skill7 starsChanged 6 months agoMITView on GitHub

Daily-style deck read — active and impeded cards, closures within the last 24h, cards parked on a human decision gate. Read-only, never mutates state. AUTO-INVOKE on "where do we stand", "what's stuck", "standup", "daily check", or "what happened since yesterday".

57 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use close-day, meeting-prep and 9 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • vantageos-agencySkill

    close-day

    End-of-day routine: close tasks with their time, leave the repository and the disk clean, write the diary, harvest friction, and hand over to tomorrow. Use this skill whenever the user says "close day", "end of day", "fin de journée", "bonne nuit", "wrap up", "call it a day", "close session", "daily close", or mentions ending their work session -- even if they don't say "close-day" explicitly.

    Also in 3 projects that use this

  • OpenAISkill

    meeting-prep

    Use when creating Public Equity Investing meeting prep briefs. Do not use for private diligence, IB, FP&A, legal, or scheduling-only tasks.

    Also in 5 projects that use this

  • jasonmichaelbell78-creatorSkill

    todo

    Cross-session todo management with interactive menu, JSONL-backed persistence, and AI-powered context capture. Manages ideas, tasks, and insights that emerge during sessions so they survive across sessions. Independent of GSD.

    Also in 3 projects that use this

  • BugRogerSkill

    project

    Project lifecycle management. Dispatches to Ada.

    Also in 3 projects that use this

  • asterkinSkill

    adr

    Create Architecture Decision Records (ADRs) following project standards. Guides template usage, process, and best practices for documenting architecturally significant decisions.

    Also in 4 projects that use this

  • VercelSkill

    deploy

    Deploy applications to Vercel. Use when the user says "deploy", "deploy to Vercel", "push to production", "deploy my app", or "go live".

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 53 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If any ! block below shows goc: command not found, Permission for this action has been denied, or no such file or directory: …

    line 17
  2. Very specific:

    If the first Context block's branch line shows [behind N], local is behind upstream: say so at the top of your report and treat every …

    line 19
  3. Distinctive:

    If none match, report "Nothing closed in the last 24h." When the count is large (a batch sync), summarize by theme rather than listing …

    line 90

“Common” is measured: kept by at least 6 in 10 of the 53 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

When to invoke

Invoke when the user says "what's up", "where do we stand", "what's blocked", "what's stuck", "daily check", "standup", "morning check", or "what happened since yesterday".

Preflight

If any ! block below shows goc: command not found, Permission for this action has been denied, or no such file or directory: .game-of-cards/deck/, stop and invoke Skill(kickoff) first. Kickoff finds the missing setup step (CLI not installed, deck not scaffolded); a denied goc call goes to Skill(claude-kickoff) on Claude Code. Re-invoke this skill only after kickoff completes.

If the first Context block's branch line shows [behind N], local is behind upstream: say so at the top of your report and treat every section below as possibly stale — closures and status changes pushed to the remote will not appear until the user pulls.

Context

!git fetch --quiet 2>/dev/null; git status -sb 2>/dev/null | head -1

!b=.claude/skills/_goc-bootstrap.sh; if [ -f $b ]; then sh $b --status active -v; else goc --status active -v; fi 2>&1 || true

!b=.claude/skills/_goc-bootstrap.sh; if [ -f $b ]; then sh $b --waiting -v; else goc --waiting -v; fi 2>&1 | head -60

!b=.claude/skills/_goc-bootstrap.sh; if [ -f $b ]; then sh $b --status open --json; else goc --status open --json; fi 2>&1 | head -60

Standup

Scrum's Daily Scrum (Schwaber & Sutherland) condensed to three questions: what's running, what's stuck, what's new? This skill collapses a remote-sync check, scan-deck, and the deck's structured closure record (closed_at) into one read.

Read-only — this skill never mutates card state (the git fetch in Context updates only remote-tracking refs, not your working tree or any card). It is safe to invoke at any time, as often as needed.

Section 1 — In flight

For each active card (shown above): report title, summary, and who claimed it (worker field). If the card has been active longer than 48h without a log.md update, flag it as potentially stalled.

Section 2 — Impeded (waiting overlay)

For each card carrying a waiting_on overlay (shown above): report title, the waiting_on reason, the waiting_until date if any, and the body's ## Waiting section (or the most recent log.md entry if no section exists). One line per card. A card may appear here even while status: active — the overlay is orthogonal to the progress status, which is why the Context block above asks for the engine's own impediment view, goc --waiting: every status, minus terminal and draft cards (engine.live_impeded), never just the open queue.

Section 3 — Closed since yesterday

List cards whose engine-maintained closed_at falls within the last 24 hours. Read the structured record — not file mtime. A find … -newer … -mmin scan is git-blind: a pull/merge/clone writes the deck directory and every log.md at one instant, giving them identical mtimes, so a strict -newer test matches nothing and standup falsely reports "Nothing closed" right after a sync lands the day's work. closed_at is the timestamp the engine sets on close, so it is immune to that and uniformly covers done / disproved / superseded.

goc --json --closed-since 24h --slim 2>/dev/null | python3 -c "
import json, sys
cards = json.load(sys.stdin)
for c in cards:
    print(f\"{c['title']}: {c['status']}\")
print(f\"({len(cards)} closed in last 24h)\" if cards else 'Nothing closed in the last 24h.')
" 2>/dev/null || true

--closed-since 24h does the date-window filter inside the engine (reading the structured closed_at record, not file mtime — a pull/merge/clone writes every log.md at one instant and a -newer-mtime scan falsely reports "Nothing closed") and lists the closures most recently closed first, so the snippet prints them as they come. --slim strips body/large fields so the Context block ships kilobytes instead of the multi-hundred-KB full-deck JSON. The title-by-status output is enough for the standup format; expand the window or drop --slim if you need summaries.

If none match, report "Nothing closed in the last 24h." When the count is large (a batch sync), summarize by theme rather than listing every line.

Section 4 — Waiting on you

Surface all cards with human_gate: decision or human_gate: session, oldest first. These are parked cards where the agent hit the Andon cord and a human must lower the gate to unblock autonomous work.

goc --json --status all 2>/dev/null | \
  python3 -c "
import json, sys
cards = json.load(sys.stdin)
waiting = [c for c in cards
           if c.get('human_gate') in ('decision', 'session')
           and c['status'] not in ('done', 'disproved', 'superseded')
           and not c.get('draft')]
waiting.sort(key=lambda c: c.get('created', ''))
for c in waiting:
    print(f\"{c['title']} [{c['human_gate']}]: {c.get('summary', '(no summary)')[:80]}\")
" 2>/dev/null || true

Like the overlay in Section 2, the gate is orthogonal to the progress status: a card may appear here while status: active, because the ordinary Andon-cord shape is claim → work → hit a judgement call → raise the gate, which leaves the card parked at active. So the query spans --status all rather than the open queue, then drops by hand the two classes a widened scope would otherwise add — closing never lowers a gate, so a terminal card can carry a stale one, and goc new gates at decision by default, so every unauthored draft scaffold is gated.

For each: show title, gate type, and the ## Decision required section header if present. The human resolves these via Skill(decide-card).

Section 5 — Next up

Show the top 3 open human_gate: none cards by value score (the cards Skill(pull-card) would pick next), as a forward look.

goc --ready 2>/dev/null | grep -v '^ACTIVE:' | head -5 || true

--ready is the engine's pull predicate (status: open ∧ not a draft ∧ human_gate: none ∧ no active impediment overlay) — the same selection Skill(pull-card) makes. Bare goc is only the first of those four conjuncts, and because the queue is value-sorted while gated epics carry the top values, it surfaces the least pullable cards in the deck. When --ready matches nothing, report that instead of omitting the section: an empty pull queue is the day's headline, and Section 4 is its cause.

The grep -v drops the engine's ACTIVE: soft-lock banner, which goes to stdout and appears only when some card is claimed. Left in, it spends one of the five budgeted lines on chrome, so the block delivers two rows instead of the three promised above — keep it filtered, and keep the budget at 5. Nothing is lost: Section 1 already renders those active cards in full. The No cards match line does not start with ACTIVE:, so the empty-queue report still survives the filter.

Output format

⚠️ Behind upstream by <N> — pull first; view may be stale.   (only if the branch line shows [behind N])

## In flight
- <title> (claimed by <worker>, <N>h): <one-line summary>

## Impeded
- <title> [waiting_on: <reason>{ until <date>}]: <one-line reason>

## Closed since yesterday
- <title>: <status> — <one-line what-changed>
  (or, for a large batch: "<N> closed — <theme>: a, b, c; <theme>: d, e")

## Waiting on you
- <title> [decision]: <decision question, ≤80 chars>

## Next up
- <title> (value: <N>): <one-line summary>

Sections with no entries are omitted — except "Next up", which reports "Nothing pullable" rather than vanishing, because an empty pull queue is a finding, not an absence. Total output ≤ 40 lines. The sync warning, if present, always goes first — never bury a stale-data caveat.

Cross-references

  • Skill(decide-card) — lower a decision/session gate so agents can resume.
  • Skill(scan-deck) — deeper browse of the deck (filtered queues, kanban board).
  • Skill(pull-card) — claim and work the top open card.
  • Skill(retrospective) — backwards analysis of recently closed cards.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.