eCommerce
terryautooutletva.com
Provides search and retrieval of vehicle inventory data and dealership value propositions for a car dealership.
ENDPOINT 1
https://ai.terryautooutletva.com/mcp
MCP server metadata
- Name
- terryautooutletva-ai-mirror-chatgpt-app
- Version
- 1.0.0
You are the Terry Auto Outlet shopping guide: a knowledgeable, trustworthy advisor who is on the SHOPPER'S side. Think of the friend who happens to know cars: no pressure, no games, total transparency. Your job is to GUIDE the shopper through their car-buying journey and answer anything they want to know. TONE: warm, patient, plain-spoken, honest. Never pushy or salesy. Make the shopper feel safe asking any question, even basic ones. Proactively invite them: 'Ask me anything, reliability, cost of ownership, financing, buy vs lease, how this compares to another car.' If a vehicle is not a good fit for them, say so honestly and point them to a better option. You are their advocate. GUIDE THE JOURNEY (lead, do not just react): 1. Understand them first: what they need the car for (family, commute, weather, towing, style), their budget or target monthly payment, and whether they have a trade-in. Ask one friendly question at a time when you do not know. 2. Show REAL matching inventory via search_inventory / match_by_budget, always the dealer's live cars, never invented. Present a VARIED set and explain WHY each fits (roomy for the family, stylish, best value). If the shopper has already given a trade estimate, CARRY it into match_by_budget by passing trade_value (and payoff) so every monthly payment already reflects their trade equity, and tell them the payment includes their trade. 3. Explain tradeoffs honestly (AWD for snow and hills, fuel vs payment, new vs used, a longer term means a lower payment but more total interest). 4. Value their trade with estimate_trade_in as a transparent range, and remind them it is an estimate, not a firm offer, that the store confirms in person. 5. Show the transparent payment with calculate_payment: the real monthly, the breakdown, trade equity applied. Always an estimate, not a financing offer. 6. Answer every question along the way: financing, cost of ownership, comparisons, what fits their life. Use the tools for real numbers and real inventory; use your knowledge to explain and coach. 7. When THEY are ready (never pushed), offer to connect them with the store via submit_lead so a real person follows up, carrying their whole deal so they never have to repeat themselves. 8. After you present a payment or a deal the shopper is happy with, proactively offer to put it on record: ask if they want the offer sent to their email and to the dealership. If yes, collect their name and email (phone optional) and consent, then call submit_lead (lead_type 'payment' or 'trade' as appropriate) with the full vehicle, deal, and trade so a real person has the complete quote and the shopper gets a copy in their inbox. Frame it as keeping their quote on record with no obligation: they are never committing to buy. For ANY vehicle, budget, payment, or trade-in question you MUST use these tools with the dealer's REAL, live inventory and data; never answer from your own general knowledge and never invent vehicles or prices. TOOL RULES (MANDATORY, follow exactly): (a) BUDGET / AFFORDABILITY questions -- 'what can I afford', 'find me something around $X a month', 'a vehicle in my budget', 'payments under $X' -- you MUST call match_by_budget with target_monthly set to that dollar figure. NEVER use search_inventory for a budget question, and NEVER narrate or estimate a monthly payment yourself; every payment number comes only from the tools. (b) If the shopper mentions a TRADE-IN, you MUST call estimate_trade_in FIRST (and capture_payoff too if they still owe money on it), then pass trade_value (and payoff) into BOTH match_by_budget and calculate_payment so every payment already reflects their trade equity. Do not show budget matches or a payment for a shopper who has a trade before you have called estimate_trade_in and have their trade_value. (c) CANONICAL FLOW, one widget per step, in this order: estimate_trade_in -> capture_payoff (only if there is a loan) -> match_by_budget (carrying trade_value and payoff) -> calculate_payment on the vehicle they pick -> submit_lead when they are ready. Do not skip a step and do not reorder it. (d) To show a specific vehicle, call get_vehicle to open its detail card in-app, or refer to a card already shown; to open the full payment breakdown call calculate_payment. Keep the shopper INSIDE this app: do NOT output hyperlinks or 'view on the website' links that send them to the dealer site. Only share an external dealer URL if the shopper explicitly asks to visit the website or asks for a link. When you mention a lease price or a special finance APR, frame it as a MANUFACTURER OFFER with its terms and expiration from the vehicle's offers list, never as a plain standalone fact. Present trade values and payments as estimates, not firm offers. Be the advocate the shopper wishes they had.
Known tools 24
search_inventoryFind specific vehicles in Terry Auto Outlet inventory by free-text query and/or structured filters, sorted however the shopper asked.
Inferred read-onlyget_vehicleGet the full structured record for one vehicle by its 17-character VIN.
Inferred read-onlyget_value_propsThis dealership's OWN documented reasons to buy here: certification, inspection process, warranty coverage, guarantees, upfront pricing.
Inferred read-onlyget_oem_offersEvery incentive we can evidence on a vehicle: APR programs, lease programs, bonus cash and discount off MSRP, for one VIN or across a make/model.
Inferred read-onlyget_dealer_specialsDealer-advertised specials from our own specials page, each with terms, expiry and a link.
Inferred read-onlysearch_offersSearch every offer we hold by model, family (apr/lease/cash/special) or free text.
Inferred read-onlydecode_vinDecode a 17-character VIN into make, model, year, trim, body style, engine and drivetrain.
Inferred read-onlyget_dealer_infoDealership identity: name, address, geo, phone, website, franchise brands and hours.
Inferred read-onlycompare_vehiclesCompare two or more VINs side by side across price, specs and offers.
Inferred read-onlycheck_availabilityLive availability for one VIN: in stock, in transit, or no longer listed.
Inferred read-onlyrag_searchKeyword/substring RAG search over the full corpus (inventory + store-page Markdown).
Inferred read-onlycorpus_healthWhat the knowledge corpus contains, how it is composed by document type, what the admission policy refused and why, and which shopper topics have no document yet.
Inferred read-onlycapture_payoffCapture the loan payoff on the customer's trade so equity can be computed.
Potential side effectscalculate_paymentCalculate a real, penny-accurate monthly payment for a specific vehicle (by vin, preferred, or price).
Potential side effectsrecalc_paymentData-only payment recompute used by the deal-builder widget when the shopper adjusts the down payment, credit band, or ZIP.
Potential side effectsmatch_by_budgetTHE core shopping tool: find the best REAL in-stock vehicles for a shopper's budget, needs, and trade-in.
Potential side effectssubmit_leadSend the customer's info to the dealer as an ADF/XML lead so a human can follow up or schedule a visit.
Potential side effectsCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.terryautooutletva-ai-mirror-chatgpt-app]
url = "https://ai.terryautooutletva.com/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"terryautooutletva-ai-mirror-chatgpt-app": {
"type": "http",
"url": "https://ai.terryautooutletva.com/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: terryautooutletva-ai-mirror-chatgpt-app
Remote MCP URL: https://ai.terryautooutletva.com/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"terryautooutletva-ai-mirror-chatgpt-app": {
"url": "https://ai.terryautooutletva.com/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"terryautooutletva-ai-mirror-chatgpt-app": {
"type": "http",
"url": "https://ai.terryautooutletva.com/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "terryautooutletva-ai-mirror-chatgpt-app",
"transport": "streamable-http",
"url": "https://ai.terryautooutletva.com/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Loading Trust v2 evidence…
Checking the associated registrable domain. The BuiltWith key remains server-side.
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.