# Automotive AI Skills

> Open-source collection of 20 AI skills (structured prompt frameworks) for automotive diagnostics, maintenance, repair, customization, EVs, motorcycles, service intake, estimate review, sales listings, and recall checks. Platform-agnostic — works with Claude, ChatGPT, or any capable LLM. Content available in English and Japanese (all skills) and Spanish (core skills). MIT licensed.

Key facts:
- Each skill defines capabilities, ready-to-use example prompts, a recommended workflow, and honest limitations.
- Safety-first design: torque specs and capacities are always ranges to verify against factory documentation; brakes, steering, SRS, fuel, and EV high-voltage work is deferred to professionals.
- The project never provides guidance on defeating emissions controls, safety systems, odometers, or immobilizers.
- Machine-readable metadata: each skill folder has a skill.yaml; known issues have a structured index.yaml.

## Skills

- [Mechanic Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/mechanic-assistant/): repair procedures, torque spec guidance, tools, system explanations
- [OBD-II Diagnostic Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/obd2-diagnostic-assistant/): trouble codes (DTC), freeze frame, live data interpretation
- [Vehicle Maintenance Planner](https://shoka-jp.github.io/automotive-ai-skills/skills/vehicle-maintenance-planner/): preventive maintenance schedules, severe-service adjustments, catch-up plans
- [Used Car Inspector](https://shoka-jp.github.io/automotive-ai-skills/skills/used-car-inspector/): pre-purchase checklists, red flags, negotiation support
- [Vehicle Troubleshooting](https://shoka-jp.github.io/automotive-ai-skills/skills/vehicle-troubleshooting/): symptom-first diagnosis — noises, vibrations, smells, leaks, warning lights
- [Vehicle Customization Advisor](https://shoka-jp.github.io/automotive-ai-skills/skills/vehicle-customization-advisor/): modification planning, fitment math, build order, legality flags
- [EV Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/ev-assistant/): charging strategy, battery health, range, EV-specific maintenance
- [Motorcycle Mechanic](https://shoka-jp.github.io/automotive-ai-skills/skills/motorcycle-mechanic/): chains, carbs and EFI, valve clearances, suspension setup
- [Diesel Specialist](https://shoka-jp.github.io/automotive-ai-skills/skills/diesel-specialist/): common rail, DPF/EGR/SCR (AdBlue), glow plugs, turbos
- [Classic Car Restorer](https://shoka-jp.github.io/automotive-ai-skills/skills/classic-car-restorer/): points ignition, carburetors, rust assessment, parts sourcing
- [Fleet Manager](https://shoka-jp.github.io/automotive-ai-skills/skills/fleet-manager/): multi-vehicle scheduling, cost tracking, repair-vs-replace analysis
- [Tire & Wheel Advisor](https://shoka-jp.github.io/automotive-ai-skills/skills/tire-wheel-advisor/): sizing math, fitment, seasonal strategy, wear diagnosis, TPMS
- [Commercial Vehicle & Truck Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/commercial-vehicle-assistant/): air brakes, pre-trip inspections, load securing
- [Body & Paint Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/body-paint-assistant/): dent/scratch assessment, repair paths, paint matching, DIY limits
- [Towing & Trailer Advisor](https://shoka-jp.github.io/automotive-ai-skills/skills/towing-trailer-advisor/): hitches, tongue weight, trailer wiring and brakes, sway prevention
- [Car Audio & Electronics Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/car-audio-electronics/): head units, amps, dash cams, 12V wiring discipline, CAN-bus limits

- [Service Intake Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/service-intake-assistant/): Turn customer symptoms into an evidence-labeled intake note, focused follow-up questions, and a technician handoff.
- [Repair Estimate Reviewer](https://shoka-jp.github.io/automotive-ai-skills/skills/repair-estimate-reviewer/): Compare repair estimates by scope, diagnostic evidence, parts, labor, arithmetic, and questions to resolve before authorization.
- [Used Car Listing Assistant](https://shoka-jp.github.io/automotive-ai-skills/skills/used-car-listing-assistant/): Draft used-vehicle listings from documented specifications, condition disclosures, and an explicit price breakdown without unsupported claims.
- [Vehicle Recall Checker](https://shoka-jp.github.io/automotive-ai-skills/skills/vehicle-recall-checker/): Guide official recall lookups, separate model-level notices from vehicle applicability, and document completion evidence or unknown status.

## Guides

- [Hardware-free MCP demo](https://shoka-jp.github.io/automotive-ai-skills/docs/try-obd2-mcp.html): source-build commands, captured simulated output, test provenance and limitations; no AI subscription required
- [MCPデモ（日本語）](https://shoka-jp.github.io/automotive-ai-skills/docs/try-obd2-mcp.ja.html): 車なしで故障コードとセンサー値を読む

- [Getting Started](https://shoka-jp.github.io/automotive-ai-skills/docs/getting-started.html): five-minute walkthrough from finding the repo to a productive session
- [FAQ](https://shoka-jp.github.io/automotive-ai-skills/docs/faq.html): trust, accuracy, platform, and usage questions answered directly
- [Safety Disclaimer](https://shoka-jp.github.io/automotive-ai-skills/docs/safety-disclaimer.html): verification rules and professional-territory boundaries
- [Skill Authoring Guide](https://shoka-jp.github.io/automotive-ai-skills/docs/skill-authoring-guide.html): required structure and quality bar for new skills
- [Japan Guide (日本語)](https://shoka-jp.github.io/automotive-ai-skills/docs/ja/japan-guide.html): US-to-Japan reading table, kei cars, terminology glossary
- [Shaken Checklist (車検, 日本語)](https://shoka-jp.github.io/automotive-ai-skills/docs/ja/shaken-checklist.html): pre-inspection self-check for Japan's vehicle inspection

## Data

- [Known-Issues Database](https://shoka-jp.github.io/automotive-ai-skills/data/known-issues/): community-contributed model-specific issues — 17 entries covering BMW, Toyota, Subaru, Honda, Nissan, Mazda, Suzuki, and Daihatsu (engines, transmissions, fuel systems, brakes, and A/C) and growing, with a machine-readable index.yaml
- [Example Conversations](https://shoka-jp.github.io/automotive-ai-skills/examples/): illustrative teaching conversations, not verified customer results — P0420 diagnosis, used-car inspection, maintenance catch-up plan

## Optional

- [Integrations](https://shoka-jp.github.io/automotive-ai-skills/integrations/): Claude Code plugin (/plugin marketplace add shoka-jp/automotive-ai-skills), OpenAI Codex guide (AGENTS.md), Claude Project template, and a combined Garage Assistant system prompt
- [Japanese README (日本語)](https://shoka-jp.github.io/automotive-ai-skills/README.ja.html)
- [Spanish README (Español)](https://shoka-jp.github.io/automotive-ai-skills/README.es.html)
- [GitHub repository](https://github.com/shoka-jp/automotive-ai-skills)
- [obd2-mcp](https://github.com/shoka-jp/obd2-mcp): companion MCP server that reads real vehicles (DTCs, live data, freeze frame, VIN) via ELM327 adapters; mock vehicle included

## Interactive entry point

- [Vehicle Brief Builder](https://shoka-jp.github.io/automotive-ai-skills/tools/vehicle-brief.html): browser-only source-linked checklist and AI prompt builder; eight Japan-market profiles and six tasks; no AI inference or official eligibility lookup
- [車種別チェックメーカー](https://shoka-jp.github.io/automotive-ai-skills/tools/vehicle-brief.ja.html): 日本語版。入力はページ内だけで処理し、質問文のコピー・保存・印刷が可能
- [Quality and evidence](https://shoka-jp.github.io/automotive-ai-skills/docs/quality.html): deterministic and browser checks, answer-review cases, limitations, and evidence contribution loop
- [Model-reference routing](https://shoka-jp.github.io/automotive-ai-skills/docs/model-reference-routing.html): shared CLI/browser lookup with exact profile, code, production-year, and market boundaries