Classify a cliente issue, atribuir prioridade, route it, e draft o primeiro resposta. — Claude Skill

name: ticket-triage description: triagem e priorizar a suporte ticket ou cliente issue. usar when a novo ticket comes in e precisa de categorization, assigning P1-P4 prioridade, deciding which equipa deve handle it, ou checking whether it's a duplicate ou known issue antes routing. argument-hint: "<ticket ou issue description>"

/ticket-triage

If você see unfamiliar placeholders ou precisar de para verificar which tools are connected, see CONNECTORS.md.

Categorize, priorizar, e route an incoming suporte ticket ou cliente issue. Produces a structured triagem assessment com a suggested initial resposta.

Usage

/ticket-triage <ticket text, cliente mensagem, ou issue description>

Examples:

• ¤KEEP0¤
• ¤KEEP0¤
• ¤KEEP0¤
• ¤KEEP0¤

workflow

1. Parse o Issue

ler o input e extrair:

• Core problem: What is o cliente actually experiencing?
• Symptoms: What specific behavior ou error are they seeing?
• cliente contexto: Who is this? qualquer conta details, plano level, ou history disponível?
• urgência sinais: Are they blocked? Is this production? How many utilizadores affected?
• Emotional declarar: Frustrated, confused, matter-de-fact, escalating?

2. Categorize e priorizar

Using o category taxonomy e prioridade framework below:

• atribuir a primary category (bug, how-para, feature pedido, billing, conta, integration, security, dados, performance) e an opcional secondary category
• atribuir a prioridade (P1–P4) based on impacto e urgência
• Identify o produto area o issue mapeia para

3. verificar para Duplicates e Known Issues

antes routing, verificar disponível fontes:

• ~~suporte platform: pesquisar para similar open ou recently resolved tickets
• ~~knowledge base: verificar para known issues ou existing documentation
• ~~project tracker: verificar if there's an existing bug relatório ou feature pedido

Apply o duplicate detection processo below.

4. Determine Routing

Using o routing rules below, recomendar which equipa ou fila deve handle this based on category e complexity.

5. gerar triagem Output

## triagem: [One-line issue summary]

**Category:** [Primary] / [Secondary if applicable]
**prioridade:** [P1-P4] — [brief justification]
**produto area:** [Area/team]

### Issue Summary
[2-3 sentence summary de what o cliente is experiencing]

### Key Details
- **cliente:** [nomear/account if known]
- **impacto:** [Who e what is affected]
- **Workaround:** [disponível / não disponível / Unknown]
- **Related tickets:** [Links para similar issues if found]
- **Known issue:** [Yes — link / No / Checking]

### recomendação de encaminhamento
**Route para:** [equipa ou fila]
**Why:** [brief reasoning]

### Suggested Initial resposta
[Draft primeiro resposta para o cliente — acknowledge o issue,
set expectations, provide workaround if disponível.
usar o auto-resposta templates below as a starting point.]

### notas internas
- [qualquer additional contexto para o agent picking this up]
- [Reproduction hints if it's a bug]
- [escalamento gatilhos para watch para]

6. Offer Próximos passos

depois presenting o triagem:

• "Want me para draft a full resposta para o cliente?"
• "deve I pesquisar para more contexto on this issue?"
• "Want me para verificar if this is a known bug in o tracker?"
• "deve I escalate this? I can package it com /customer-escalation."

Category Taxonomy

atribuir cada ticket a primary category e optionally a secondary category:

Category Determination Tips

• If o cliente relatórios both a bug e a feature pedido, o bug is primary
• If they can't log in due para a bug, category is bug (não conta) — root cause drives o category
• "It used para work e now it doesn't" = bug
• "I want it para work differently" = Feature pedido
• "How do I make it work?" = How-para
• When in doubt, lean toward bug — it's better para investigate than dismiss

prioridade Framework

P1 — Critical

Criteria: Production system down, dados loss ou corruption, security breach, todos ou most utilizadores affected.

• o cliente cannot usar o produto at todos
• dados is being lost, corrupted, ou exposed
• A security incidente is in progress
• o issue is worsening ou expanding in scope

SLA expectation: Respond dentro de 1 hour. Continuous work until resolved ou mitigated. Updates cada 1-2 hours.

P2 — High

Criteria: Major feature broken, significant workflow blocked, many utilizadores affected, no workaround.

• A core workflow is broken but o produto is partially usable
• Multiple utilizadores are affected ou a key conta is impacted
• o issue is blocking time-sensitive work
• No reasonable workaround exists

SLA expectation: Respond dentro de 4 hours. Active investigation same day. Updates cada 4 hours.

P3 — Medium

Criteria: Feature partially broken, workaround disponível, single utilizador ou small equipa affected.

• A feature isn't working correctly but a workaround exists
• o issue is inconvenient but não blocking critical work
• A single utilizador ou small equipa is affected
• o cliente is não escalating urgently

SLA expectation: Respond dentro de 1 business day. resolução ou update dentro de 3 business days.

P4 — Low

Criteria: Minor inconvenience, cosmetic issue, general question, feature pedido.

• Cosmetic ou UI issues that don't affect functionality
• Feature pedidos e enhancement ideas
• General questions ou how-para inquiries
• Issues com simple, documented solutions

SLA expectation: Respond dentro de 2 business days. resolução at normal pace.

prioridade escalamento gatilhos

Automatically bump prioridade up when:

• cliente has been waiting longer than o SLA allows
• Multiple clientes relatório o same issue (pattern detected)
• o cliente explicitly escalates ou mentions executivo involvement
• A workaround that was in place stops working
• o issue expands in scope (more utilizadores, more dados, novo symptoms)

Routing Rules

Route tickets based on category e complexity:

Duplicate Detection

antes creating a novo ticket ou routing, verificar para duplicates:

1. pesquisar by symptom: Look para tickets com similar error mensagens ou descriptions
2. pesquisar by cliente: verificar if this cliente has an open ticket para o same issue
3. pesquisar by produto area: Look para recent tickets in o same feature area
4. verificar known issues: comparar against documented known issues

If a duplicate is found:

• Link o novo ticket para o existing one
• Notify o cliente that this is a known issue being tracked
• Add qualquer novo information a partir do novo relatório para o existing ticket
• Bump prioridade if o novo relatório adds urgência (more clientes affected, etc.)

Auto-resposta templates by Category

bug — Initial resposta

Thank você para reporting this. I consegue ver how [specific impacto]
would be disruptive para o seu work.

I've logged this as a [prioridade] issue e our equipa is
investigating. [If workaround exists: "In o meantime, você
can [workaround]."]

I'll update você dentro de [SLA timeframe] com what we encontrar.

How-para — Initial resposta

Great question! [Direct answer ou link para documentation]

[If more complex: "Let me walk você through o steps:"]
[Steps ou guidance]

Let me know if that ajuda, ou if você have qualquer follow-up
questions.

Feature pedido — Initial resposta

Thank você para this suggestion — I consegue ver why [capability]
would be valuable para o seu workflow.

I've documented this e shared it com our produto equipa.
While I can't commit para a specific timeline, o seu feedback
directly informs our roadmap prioridades.

[If alternative exists: "In o meantime, você might encontrar
[alternative] helpful para achieving something similar."]

Billing — Initial resposta

I understand billing issues precisar de prompt attention. Let me
look em this para você.

[If straightforward: resolução details]
[If complex: "I'm reviewing o seu conta now e will have
an answer para você dentro de [timeframe]."]

Security — Initial resposta

Thank você para flagging this — we take security concerns
seriously e are reviewing this immediately.

I've escalated this para our security equipa para investigation.
We'll follow up com você dentro de [timeframe] com our findings.

[If ação is needed: "In o meantime, we recomendar
[protective ação]."]

triagem Best Practices

1. ler o full ticket antes categorizing — contexto in later mensagens often changes o assessment
2. Categorize by root cause, não just o symptom described
3. When in doubt on prioridade, err on o side de higher — it's easier para de-escalate than para recover a partir da missed SLA
4. Always verificar para duplicates e known issues antes routing
5. Write notas internas that ajudar o próximo person pick up contexto quickly
6. incluir what você've already checked ou ruled out para avoid duplicate investigation
7. assinalar patterns — if você're seeing o same issue repeatedly, escalate o pattern even if individual tickets are low prioridade