Recommend a car variant or model based on user needs, research data, and structured questionnaire.
Use this skill when the user asks for a recommendation on which car or variant to buy. Combine research data with a structured questionnaire to produce evidence-based suggestions.
Use the questions in research/variant-suggestion-questionnaire.md.
Use the compact version (~10 key questions) for a fast consultation.
Best for: users who already have a shortlist, users who want a quick answer, follow-up suggestions.
Use the full questionnaire (10+ sections, 90+ questions) for a deep buyer profile.
Best for: users who are undecided, first-time buyers, users comparing across multiple models.
Default to Quick mode unless the user explicitly requests a thorough or comprehensive analysis.
This skill supports two suggestion types:
Help the user choose between different car models.
Uses Section 0 (Car Selection) and Cross-Model Decision Filters from the questionnaire.
Help the user choose between variants of the same model.
Uses Sections 1-9 of the questionnaire.
Support every recommendation with specific evidence from research files in the research/ folder.
If research data for a car is insufficient or missing, trigger a Research workflow first before making recommendations.
Do not make unsupported claims. Every "this variant is better because..." must link to a specific data point from research.
After collecting answers, produce:
The variant or model most aligned with the user's stated needs.
The option with the best price-to-feature ratio given the user's priorities.
The lowest-cost option that still meets the user's minimum requirements.
The top-tier option that makes sense if the user is willing to stretch budget for maximum features.
Clear warnings for each recommendation: "Avoid this if you..."