Local Auto Parts Quote Assistant

Problem

Auto parts quoting often breaks down when local Excel data is shared by multiple users. A single part number may have high-quality, aftermarket, 4S, and OEM prices, while names, vehicle models, and notes are split across worksheets. Manual lookups can miss fields, mix up price tiers, or reuse stale results. This skill targets fixed quote rules and uses ~/Desktop/报价系统.xlsx as the single source of truth, returning part-number based Markdown quote results that can be reviewed line by line.

How It Works

• Real-time reloads: runs quote_v4.py to read the 总数据 and 价格 sheets on every query, without relying on conversation cache.
• Part number matching: checks both the main catalog and price sheet; price-sheet-only hits can still return a price, while name and vehicle model may be empty.
• Price tier separation: no suffix or -ZC maps to high-quality pricing, -B/-A maps to aftermarket pricing, total-data column 10 maps to 4S pricing, and adjusted price is multiplied by 1.03 for OEM pricing.
• Result output: first shows statistics, data status, and warnings, then outputs a fixed 16-column Markdown table, and finally asks whether to export Excel.
• Export and audit: after confirmation, generates a styled desktop Excel file and records caller, input part numbers, matched count, warning count, and output path.

Boundaries

This fits teams that already maintain a local quote workbook and need fast lookup and export. It is not a rules engine or data backend: it should not invent missing prices, guess nearby parts, hand-build quote tables, or change price logic. If a part is not found, a field shows -, a warning appears, OEM price is 0, or adjusted price is empty, the user should be asked to confirm manually.