Overview
Customer support without SLA-based ticket routing, knowledge-base-assisted resolution, and automated escalation does not scale. The infrastructure behind that gate is ticket lifecycle management with priority queuing, canned-response libraries, and resolution analytics.
The customer support specialist enforces strict SLA adherence, self-service deflection, and resolution quality metrics across all customer touchpoints.
When to Use
- When implementing ticket triage systems with priority levels (P1-Critical, P2-High, P3-Medium, P4-Low).
- When designing refund request workflows with approval gates and audit trails.
- When building live-chat bots, FAQ systems, and knowledge-base deflection strategies.
- When setting up SLA tracking: response time, resolution time, customer satisfaction (CSAT).
- When designing customer onboarding sequences, welcome emails, and success checklists.
- When implementing escalation paths: Level 1 (support), Level 2 (specialists), Level 3 (engineering).
When NOT to Use
- When closing deals or managing sales pipeline (use
sales/role-sales-engineer).
- When handling marketing campaigns or content creation (use
marketing/role-growth-marketer).
- When building product features or bug fixes (use
development/role-software-engineer).
Deterministic Dispatch Matrix
| Support Objective |
Delegated Child Skill |
Primary Support Artifact |
| Ticket triage & priority routing |
sales/business-development |
Priority matrix + escalation rules |
| Refund request workflow |
sales/high-ticket-closing |
Refund policy + approval chain |
| Live chat bot setup |
automation/n8n-workflows |
Chatbot flow + fallback routing |
| Knowledge base & FAQ |
content/writing |
FAQ articles + search optimization |
| Onboarding emails |
marketing/email-marketing |
Onboarding sequence + templates |
| SLA dashboard |
sales/sales-strategy |
SLA tracker + alert config |
| Escalation playbook |
sales/sales-pipeline |
Escalation matrix + contact list |
Workflow
Phase 1: Ticket Routing & Prioritization
- Configure P1-P4 priority levels with SLA timeframes (P1: 4h, P2: 24h, P3: 48h, P4: 5d).
- Build routing rules: channel (email/chat/phone) + category → queue.
- Set up auto-assignment: round-robin, load balancing, or skill-based.
Phase 2: Resolution & Knowledge Deflection
- Create canned responses library for 80% recurring issues.
- Build searchable knowledge base; track "resolved via KB" metric.
- Implement chat deflection: pre-chat bot asks questions; routes to agent only if needed.
Phase 3: Refund & Complaint Handling
- Refund request queue with manager approval gate.
- Track reason codes: quality, pricing, features, delivery.
- Automated follow-up: refund confirmation + feedback survey.
Phase 4: Onboarding & Success
- Welcome email sequence (Day 0, Day 1, Day 3, Day 7, Day 30).
- Product usage checklist + feature recommendations.
- Customer health scoring: usage %, support tickets, NPS.
Phase 5: Analytics & SLA Improvement
- CSAT after every resolved ticket; track by rep, category, month.
- First-response time vs resolution time; identify slow queues.
- Escalation rate: >5% = process or training needed.
Anti-Rationalization Table
| Support Excuse |
Flawed Rationalization |
Required Support Discipline |
| "Customers just need to read docs" |
"We'll add more docs later" |
KB deflection rate must hit 60% before new docs |
| "Refunds are cheap" |
"Approve all under $100 automatically" |
All refunds over $50 require manager approval |
| "Agents should know everything" |
"We'll train after hiring" |
Tiered escalation with defined ownership |
Implementation Examples
Priority Matrix (YAML)
priority_levels:
P1_CRITICAL:
response_time: 4h
resolution_time: 24h
escalation: manager
channels: ["phone", "urgent-email"]
P2_HIGH:
response_time: 24h
resolution_time: 72h
escalation: senior-agent
channels: ["email", "chat"]
P3_MEDIUM:
response_time: 48h
resolution_time: 5d
escalation: team-lead
channels: ["email", "chat"]
P4_LOW:
response_time: 5d
resolution_time: 10d
escalation: queue
channels: ["email"]
Refund Approval Flow (Bash)
#!/usr/bin/env bash
# Refund approval flow
# Returns: approve, reject, escalate
AMOUNT=$(jq -r '.amount' /tmp/refund.json)
CUSTOMER_TIER=$(jq -r '.customer_tier' /tmp/refund.json)
REASON=$(jq -r '.reason' /tmp/refund.json)
TICKETS_HISTORY=$(jq '.support_tickets | length' /tmp/data.json)
if [ "$AMOUNT" -gt 500 ]; then
echo "escalate: amount exceeds $500"
exit 1
fi
if [ "$CUSTOMER_TIER" = "churn-risk" ] && [ "$AMOUNT" -lt 100 ]; then
echo "approve: retain at-risk customer"
exit 0
fi
if [ "$REASON" = "product-defect" ]; then
echo "escalate: technical review required"
exit 1
fi
echo "reject: outside policy"
exit 2
Verification Checklist
npm install -g cli-tool
import exampleLibrary from "@example/lib";
1---2name: role-customer-support3description: Use when handling customer ticket triage, complaint resolution, refund requests, onboarding assistance, live-chat support workflows, SLA management, and escalation procedures.4domain: sales5author: oyi776version: 1.0.07category: sales8tags:9- customer-support10- support11- tickets12- complaints13- refunds14license: Apache-2.015subdomain: customer-success16triggers:17- customer support role18- ticket triage19- complaint resolution20- refund request21- customer onboarding22- live chat support23- sla management24- escalation procedure25metadata:26 source: https://skillmd.com/skills/oyi77/role-customer-support27---2829## Overview3031Customer support without SLA-based ticket routing, knowledge-base-assisted resolution, and automated escalation does not scale. The infrastructure behind that gate is ticket lifecycle management with priority queuing, canned-response libraries, and resolution analytics.3233The customer support specialist enforces strict SLA adherence, self-service deflection, and resolution quality metrics across all customer touchpoints.3435## When to Use3637- When implementing ticket triage systems with priority levels (P1-Critical, P2-High, P3-Medium, P4-Low).38- When designing refund request workflows with approval gates and audit trails.39- When building live-chat bots, FAQ systems, and knowledge-base deflection strategies.40- When setting up SLA tracking: response time, resolution time, customer satisfaction (CSAT).41- When designing customer onboarding sequences, welcome emails, and success checklists.42- When implementing escalation paths: Level 1 (support), Level 2 (specialists), Level 3 (engineering).4344## When NOT to Use4546- When closing deals or managing sales pipeline (use `sales/role-sales-engineer`).47- When handling marketing campaigns or content creation (use `marketing/role-growth-marketer`).48- When building product features or bug fixes (use `development/role-software-engineer`).4950## Deterministic Dispatch Matrix5152| Support Objective | Delegated Child Skill | Primary Support Artifact |53|---|---|---|54| Ticket triage & priority routing | `sales/business-development` | Priority matrix + escalation rules |55| Refund request workflow | `sales/high-ticket-closing` | Refund policy + approval chain |56| Live chat bot setup | `automation/n8n-workflows` | Chatbot flow + fallback routing |57| Knowledge base & FAQ | `content/writing` | FAQ articles + search optimization |58| Onboarding emails | `marketing/email-marketing` | Onboarding sequence + templates |59| SLA dashboard | `sales/sales-strategy` | SLA tracker + alert config |60| Escalation playbook | `sales/sales-pipeline` | Escalation matrix + contact list |6162## Workflow6364### Phase 1: Ticket Routing & Prioritization65661. Configure P1-P4 priority levels with SLA timeframes (P1: 4h, P2: 24h, P3: 48h, P4: 5d).672. Build routing rules: channel (email/chat/phone) + category → queue.683. Set up auto-assignment: round-robin, load balancing, or skill-based.6970### Phase 2: Resolution & Knowledge Deflection71721. Create canned responses library for 80% recurring issues.732. Build searchable knowledge base; track "resolved via KB" metric.743. Implement chat deflection: pre-chat bot asks questions; routes to agent only if needed.7576### Phase 3: Refund & Complaint Handling77781. Refund request queue with manager approval gate.792. Track reason codes: quality, pricing, features, delivery.803. Automated follow-up: refund confirmation + feedback survey.8182### Phase 4: Onboarding & Success83841. Welcome email sequence (Day 0, Day 1, Day 3, Day 7, Day 30).852. Product usage checklist + feature recommendations.863. Customer health scoring: usage %, support tickets, NPS.8788### Phase 5: Analytics & SLA Improvement89901. CSAT after every resolved ticket; track by rep, category, month.912. First-response time vs resolution time; identify slow queues.923. Escalation rate: >5% = process or training needed.9394## Anti-Rationalization Table9596| Support Excuse | Flawed Rationalization | Required Support Discipline |97|---|---|---|98| "Customers just need to read docs" | "We'll add more docs later" | KB deflection rate must hit 60% before new docs |99| "Refunds are cheap" | "Approve all under $100 automatically" | All refunds over $50 require manager approval |100| "Agents should know everything" | "We'll train after hiring" | Tiered escalation with defined ownership |101102## Implementation Examples103104### Priority Matrix (YAML)105106```yaml107priority_levels:108 P1_CRITICAL:109 response_time: 4h110 resolution_time: 24h111 escalation: manager112 channels: ["phone", "urgent-email"]113 P2_HIGH:114 response_time: 24h115 resolution_time: 72h116 escalation: senior-agent117 channels: ["email", "chat"]118 P3_MEDIUM:119 response_time: 48h120 resolution_time: 5d121 escalation: team-lead122 channels: ["email", "chat"]123 P4_LOW:124 response_time: 5d125 resolution_time: 10d126 escalation: queue127 channels: ["email"]128```129130### Refund Approval Flow (Bash)131132```bash133#!/usr/bin/env bash134# Refund approval flow135# Returns: approve, reject, escalate136137AMOUNT=$(jq -r '.amount' /tmp/refund.json)138CUSTOMER_TIER=$(jq -r '.customer_tier' /tmp/refund.json)139REASON=$(jq -r '.reason' /tmp/refund.json)140TICKETS_HISTORY=$(jq '.support_tickets | length' /tmp/data.json)141142if [ "$AMOUNT" -gt 500 ]; then143 echo "escalate: amount exceeds $500"144 exit 1145fi146147if [ "$CUSTOMER_TIER" = "churn-risk" ] && [ "$AMOUNT" -lt 100 ]; then148 echo "approve: retain at-risk customer"149 exit 0150fi151152if [ "$REASON" = "product-defect" ]; then153 echo "escalate: technical review required"154 exit 1155fi156157echo "reject: outside policy"158exit 2159```160161## Verification Checklist162163- [ ] Priority matrix defined; P1 SLA = 4h response, 24h resolution164- [ ] Canned response library complete for top 50 issues165- [ ] KB deflection rate tracked; goal > 60%166- [ ] Refund policy documented; approval gate at $50167- [ ] Escalations routed correctly; 0 stuck tickets168- [ ] CSAT tracked; > 85% target169- [ ] Escalation rate < 5%; first-contact resolution > 40%170171`npm install -g cli-tool`172173import exampleLibrary from "@example/lib";