차량 유지보수 예외 등록부 Skill

---
name: fleet-maintenance-exception-register
description: Use when a delivery, logistics, or fleet office manager needs to consolidate odometer logs, repair invoices, inspection forms, driver notes, and route schedules to identify overdue or at-risk maintenance, draft a fleet exception register, group vehicles by risk level, and escalate safety or downtime decisions to a dispatcher before scheduling service.
allowed-tools: [Read, Edit]
---
# Fleet Maintenance Exception Register
## Overview
Creates a single, auditable exception register for a vehicle fleet by consolidating maintenance-relevant inputs. Identifies overdue or at-risk maintenance, assigns risk levels, prepares dispatcher escalations for safety and downtime decisions, and proposes a service scheduling plan.

3. Normalize and reconcile
3.1. Standardize units (miles vs km), date formats, and vehicle identifiers; map aliases to canonical IDs.
3.2. Deduplicate entries; prefer the most recent dated reading for mileage.
3.3. Resolve conflicts (e.g., decreasing mileage) by flagging as data issues and noting the chosen source. Do not invent values.

4. Determine due services
4.1. For each service category (e.g., oil/filter, tire rotation, brake inspection, transmission, coolant, PM levels, DOT annual, emissions), compute next-due mileage and/or date using last service data and the confirmed intervals.
4.2. If an interval is unknown, request it or mark the service as "Interval needed" and exclude from overdue calculations until provided.

6. Group by risk level
6.1. Assign overall risk: High (any Safety-Critical or >1,000 mi/>30 days overdue), Medium (Due Soon or non-critical open defects), Low (OK).
6.2. Document the rule definitions used for the risk grouping in the Assumptions log.

7. Build the Fleet Exception Register
7.1. Create one row per vehicle containing at minimum:
- Vehicle ID (and plate/VIN if available)
- Source mileage (value, date, and source document)
- Due service(s) with due mileage/date and basis (policy/OEM)
- Risk flag/level (High/Medium/Low, plus Safety-Critical and/or Downtime Risk flags)
- Recommended action (e.g., "Escalate to dispatcher for immediate pull", "Schedule next available window", "Monitor")
- Human decision field (Dispatcher/Manager decision, name, timestamp)

8. Escalate before scheduling
8.1. For High risk and Safety-Critical items, prepare a concise escalation summary per vehicle citing sources and recommended immediate actions.
8.2. Present the summary for dispatcher decision on pull-from-route, substitution, or temporary restrictions. Pause and record the decision in the human decision field.
8.3. For Downtime Risk, analyze route schedules to propose options: swap vehicles, after-hours service, split routes, or defer within policy limits. Record the decision.