Decide

---
name: decide
description: Quick decision loop for WAITING cards. Plain text list, user picks T/B/W/D per card, transitions applied.
argument-hint: "[max_count]"
---

## /decide

Show all WAITING FOR APPROVAL cards in a mouse-selectable plain text list. User responds with quick codes to move them.

### Args

| Arg | What |
|-----|------|
| *(empty)* | All WAITING cards |
| `5` | First 5 (by priority) |

---

## How it works

Runs in main session (interactive — no background agent).

### Step 1: Fetch WAITING cards

Proje anahtarini `docs/CLAUDE_JIRA.md` veya `CLAUDE.md`'den oku.

JQL: `project = {KEY} AND status = "WAITING FOR APPROVAL" ORDER BY priority DESC`
fields: summary, priority, labels — max 50

If result is saved to file (too large), parse with python3.

### Step 2: Show cards

Plain text, one card per block — easy to mouse-select:

```
 1  {KEY}-XXX  Card title here                          High  #label
    Why waiting: one sentence reason

 2  {KEY}-YYY  Another card title                       Med   #label
    Why waiting: one sentence reason
```

At the bottom show options:

```
T = To Do    B = Backlog    W = Keep waiting    D = Close

Reply examples:
  1T 2T 3B 4W 5D
  1-5T 6-10B
  all T
```

### Step 3: Apply user decisions

Parse user reply. For each card:
- T → transition To Do (id 11) + comment "Decide: moved to To Do"
- B → transition Backlog (id 51) + comment "Decide: moved to Backlog"
- W → skip (no change)
- D → transition Done (id 31) + comment "Decide: closed"

If a transition fails, call getTransitionsForJiraIssue to find the correct ID and retry.

Show summary:
```
Done: 3 To Do | 2 Backlog | 1 Waiting | 1 Closed
```

## Rules

- Max 2 lines per card (title line + why-waiting line)
- Plain text only, NO markdown tables — must be mouse-selectable
- Extract blocker reason from description (product decision? credential? dependency?)
- Wait for single-line user reply before acting
- On transition failure → auto-retry with correct transition ID

## When NOT to Use
- Tek satirlik basit soru/cevap ise
- Skill'in scope'u disindaysa
- Riskli/destructive is ise (ayri onay gerekir)

## Red Flags
- Belirsiz hedef/kabul kriteri
- Gerekli dosya/izin/secret eksik
- Ayni adim 2+ kez tekrarlandi

## Error Handling
- Gerekli kaynak yoksa → dur, blocker'i raporla
- Komut/akıs hatasi → en yakin guvenli noktadan devam et
- 3 basarisiz deneme → daha uygun skill/agent'a yonlendir

## Verification
- [ ] Beklenen cikti uretildi
- [ ] Yan etki yok (dosya/ayar)
- [ ] Gerekli log/rapor paylasildi