damage-estimating

# Damage Estimating Single dispatcher for repair-cost logic: repair estimates, EVA/Audatex total-loss damage assessments, paint costing, and ABP charge review. ## Modes | Mode | Use for | Load | | --- | --- | --- | | `repair-estimate` | Standalone desktop estimate from vehicle details and damage photos. | `references/repair-estimate.md`, `references/estimate-framework.md`, `references/examples.md` | | `eva-total-loss` | EVA/Audatex-compatible damage-assessment or total-loss PDF. | `references/eva-total-loss.md`, `references/estimate-framework.md`, `references/examples.md` | | `paint-costing` | AZT paint time/material calculation for panels, blends, substrates, and surcharges. | `references/paint-costing.md` | | `charge-review` | ABP rate and extra-charge sanity review for a repair estimate. | `references/charge-review.md` | ## Shared References - ABP rates: `src/skills/abp-rates/references/abp-2026-rates.md`. - AZT paint method: `references/azt-paint-method.md`. - OEM constraints: `src/skills/manufacturer-standards/references/oem-index.md`. - Domain terms: `src/skills/ce-domain-glossary/references/`. - House style: `src/skills/ce-house-style/`. - Salvage category decisions: `src/skills/salvage-categorisation/` when categorisation or notification obligations are in scope. ## Required Inputs - Vehicle registration, make/model, derivative where relevant, mileage if available, and damage location. - Clear damage photos or source estimate/Audatex PDF. - Required output mode and intended audience. - Any instruction constraints, target system requirements, or known total-loss/salvage context. ## Governance `repair-estimate` and `eva-total-loss` are expert outputs and remain AI-assisted drafts until a named engineer signs off. `paint-costing` and `charge-review` are assistive calculations only. No autonomous external send. ## Validation Run repository verification after skill changes: ```powershell python tools/verify_scaffold.py ``` When rendering a valuation-linked evidence pack, also run `python tools/vehicle-valuation/validate_evidence_pack.py payload.json`.