vm-ease-classifier
Skill detail with category, linked agents, and source metadata.
vm-ease-classifier
Scores individual incident tickets against the EASE framework (Eliminate, Automate, Shift Left, Empower Users) with independent 0-1 scores per lever. Applies the decision rules in references/ease-framework.md and uses ITSM category context to produce per-incident scores and cluster-level aggregate classifications. USE FOR: per-incident EASE scoring after ITSM classification and clustering. DO NOT USE FOR: clustering tickets (use vm-ticket-cluster-analyzer), temporal pattern analysis, or explaining the EASE framework to stakeholders.
Source: .github/skills/incident-analytics/vm-ease-classifier/SKILL.md
Used By Agents
Preview
View source preview (first 3000 chars)
# vm-ease-classifier
## When To Use This Skill
- Individual incident tickets need independent 0-1 EASE lever scores after clustering and ITSM classification
- An operations team needs each incident scored on all four resolution strategies (Eliminate, Automate, Shift Left, Empower Users)
- Cluster-level EASE classification is needed by aggregating per-incident scores
**Do NOT use this skill when:**
- Ticket clustering has not yet been performed (cluster first with `vm-ticket-cluster-analyzer`)
- ITSM classification has not been applied (classify first -- ITSM categories inform EASE scoring)
- The goal is to explain the EASE framework to stakeholders (the invoking agent handles that outside this skill)
## Unitary Function
**ONE RESPONSIBILITY:** Apply the EASE decision rules from `references/ease-framework.md` to each individual ticket and return independent 0-1 scores for all four EASE levers, plus a cluster-level aggregate classification.
**NOT RESPONSIBLE FOR:**
- Producing the ticket cluster array (that is `vm-ticket-cluster-analyzer`)
- ITSM domain classification (the invoking agent handles that before calling this skill)
- Temporal pattern analysis or L-3 candidacy scoring (invoking agent's workflow responsibilities)
- Defining or explaining the EASE framework to users
- Generating the final JSON report (that is `sdlc-artifact-generator`)
- Reading ticket source files
## Input
```json
{
"tickets": [
{
"ticket_id": "INC-001",
"cluster_id": "CLU-001",
"cluster_name": "Payment Gateway Failure - Checkout",
"summary": "Payment fails at checkout for Visa cards",
"normalized_text": "payment fails checkout visa cards",
"resolution_notes": "Restarted payment gateway connector",
"area": "checkout",
"itsm_category": "Application",
"itsm_subcategory": "Payment Processing",
"top_keywords": ["payment", "visa", "checkout", "402", "gateway"]
}
],
"clusters": [
{
"cluster_id": "CLU-001",
"cluster_name": "Payment Gateway Failure - Checkout",
"ticket_count": 3,
"resolution_rate": 0.67,
"resolution_notes_sample": [
"Restarted payment gateway connector",
"Escalated to vendor"
]
}
]
}
```
| Parameter | Type | Required | Description |
|---|---|---|---|
| `tickets` | array | MUST | Individual ticket records with cluster assignment, ITSM classification, and text fields |
| `clusters` | array | MUST | Cluster summary array from `vm-ticket-cluster-analyzer` for aggregate context |
## Output
```json
{
"generated_by": {
"skill": "vm-ease-classifier",
"version": "2.0.0"
},
"ticket_scores": [
{
"ticket_id": "INC-001",
"ease_eliminate_score": 0.15,
"ease_automate_score": 0.85,
"ease_shift_left_score": 0.20,
"ease_empower_score": 0.05
}
],
"cluster_classifications": [
{
"cluster_id": "CLU-001",
"ease_class": "Automate",
"ease_rationale": "Gateway failu