Integration

DealerAI integration

DealerAI is an AI sales and service assistant for car dealerships. With this API an agent can read and search the dealership's new and used inventory and customer conversations, search and save contacts, enroll contacts in messaging sequences (which then message them) or stop them, submit ADF leads, and maintain what the assistant knows: promotions, knowledge base, staff contacts, department hours, holidays and location, then retrain it. Not exposed: the duplicate chat routes, the CreditApp partner webhook, the integration keys and Slack webhook in the dealership profile, and the assistant-thread routes found only in the guides site. Authentication is HTTP Basic with the dealership's DealerAI API username and password.

Connect DealerAI How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in DealerAI.Built from the 56 DealerAI tools below, on the connection you add once.

What you can do with DealerAI

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Each badge shows the default agent permission of an endpoint (allowed: an agent calls it without asking; ask first: you are asked before an agent calls it, when the agent or chat asks permission for sensitive actions; never: no agent may call it), which every user can change endpoint by endpoint in their credential settings.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

New, 5 operations

delete_all_new_carsDelete all new carsDELETEAgents: ask firstRemoves EVERY new car from the dealership's inventory. Cannot be undone. The answer has one row per removed car.
delete_new_carDelete new carDELETEAgents: ask firstRemoves one new car from the dealership's inventory by VIN. Cannot be undone.
get_new_carGet new carGETAgents: ask firstReturns one new car of the dealership's inventory by VIN, with price, specs, finance and lease details and price history.
list_new_carsList new carsGETAgents: ask firstLists the dealership's new-car inventory, ordered by creation date, paged with Skip and Take. Each car carries price, specs, finance and lease details and price history.
upload_new_inventory_csvUpload new inventory csvPUTAgents: ask firstUploads a CSV file of new cars that REPLACES the dealership's whole new-car inventory. Needs the file and its delimiter; the answer counts the vehicles removed and added.

Personnel, 5 operations

create_personnelCreate personnelPOSTAgents: ask firstAdds a staff member or department contact (department is required). retrain_ai with train_target KnowledgeBase retrains the AI on staff and department contacts.
delete_personnelDelete personnelDELETEAgents: ask firstDeletes a staff or department contact by id (from list_personnel). Cannot be undone.
get_personnelGet personnelGETAgents: ask firstReturns one staff or department contact by id (from list_personnel).
list_personnelList personnelGETAgents: ask firstLists the staff and department contacts the AI knows: name, title, department, email, phone, extension and notes (personal data of staff).
update_personnelUpdate personnelPUTAgents: ask firstReplaces a staff or department contact entirely: a field left out is cleared. Read it with get_personnel first and send every value back with your change.

Promotion, 5 operations

create_promotionsCreate promotionsPOSTAgents: ask firstCreates one or more promotions the AI can offer customers; each needs name and description (79 characters at most each), site_url and image_url. The answer has one row per promotion with its new id or its error.
delete_all_promotionsDelete all promotionsDELETEAgents: ask firstDeletes EVERY promotion of the dealership. Cannot be undone. The answer has one row per removed promotion.
delete_promotionDelete promotionDELETEAgents: ask firstDeletes one promotion by id (from list_promotions). Cannot be undone.
get_promotionGet promotionGETAgents: ask firstReturns one promotion by id (from list_promotions).
list_promotionsList promotionsGETAgents: ask firstLists the dealership's promotions (offers the AI can mention), optionally only active or only extracted ones, paged with Skip and Take.

Used, 5 operations

delete_all_used_carsDelete all used carsDELETEAgents: ask firstRemoves EVERY used car from the dealership's inventory. Cannot be undone. The answer has one row per removed car.
delete_used_carDelete used carDELETEAgents: ask firstRemoves one used car from the dealership's inventory by VIN. Cannot be undone.
get_used_carGet used carGETAgents: ask firstReturns one used car of the dealership's inventory by VIN, with price, mileage, CARFAX, specs, finance and lease details and price history.
list_used_carsList used carsGETAgents: ask firstLists the dealership's used-car inventory, ordered by creation date, paged with Skip and Take. Each car carries price, mileage, CARFAX, specs, finance and lease details and price history.
upload_used_inventory_csvUpload used inventory csvPUTAgents: ask firstUploads a CSV file of used cars that REPLACES the dealership's whole used-car inventory. Needs the file and its delimiter; the answer counts the vehicles removed and added.

Search, 4 operations

search_carsSearch carsPOSTAgents: ask firstSearches the dealership's new and used inventory with a natural-language query (such as red SUV under 40000) and returns the matching vehicles, paged with skip and take.
search_contactsSearch contactsPOSTAgents: ask firstSearches the dealership's contacts with filters (text, name, phone, email, sequence, vehicles of interest or owned, dates), sorted and paged. Rows are personal data, in result.rows.
search_new_carsSearch new carsGETAgents: ask firstSearches the dealership's new cars by year range, make, model, colors, body and fuel type, stock status, price range, recent price drop and tags, paged with Skip and Take.
search_used_carsSearch used carsGETAgents: ask firstSearches the dealership's used cars by year range, make, model, colors, body and fuel type, stock status, price range, recent price drop and tags, paged with Skip and Take.

Holiday, 3 operations

delete_holidayDelete holidayDELETEAgents: ask firstDeletes one holiday by holiday_id (from list_holidays). Cannot be undone. The answer lists the remaining holidays and says whether the AI was retrained.
list_holidaysList holidaysGETAgents: ask firstLists the dealership's holidays with their date and special hours per hours type, optionally filtered by holiday id, date range or text.
save_holidaysSave holidaysPUTAgents: ask firstCreates or updates holidays (pass holiday_id from list_holidays to update one) with their hours per hours type. Changes what the AI tells customers; the answer says whether the AI was retrained.

Bulk, 2 operations

bulk_enroll_contactsBulk enroll contactsPOSTAgents: ask firstEnrolls existing contacts (by contact_id) in a sequence: the sequence then MESSAGES each of them on its channels from next_message_date_time_utc on. Returns one result per contact.
bulk_stop_enrollmentsBulk stop enrollmentsPOSTAgents: ask firstStops enrollments in a sequence by enrollment id and/or contact id, cancelling their scheduled messages. One result per id; an enrollment that had already ended reports success.

Conversation, 2 operations

get_conversationGet conversationGETAgents: ask firstReturns one conversation with the customer's contact info (personal data), status and counts; set includeMessages to true to also get its messages.
list_conversationsList conversationsGETAgents: ask firstLists the dealership's customer conversations (SMS, email, webchat) with filters on channel, status and dates, ordered by last change. Rows carry the customer's name, email and phone (personal data).

Extract, 2 operations

extract_knowledge_base_entriesExtract knowledge base entriesPOSTAgents: ask firstReads a public web page and returns candidate knowledge base entries drawn from it. It saves nothing: keep the entries you want with save_knowledge_base_entries.
extract_promotionExtract promotionPOSTAgents: ask firstHas DealerAI's AI read a promotion page (markdown preferred, else HTML) and save it as a promotion. Returns the promotion, or no data when an existing one is left untouched (see overwrite and force).

Hour, 2 operations

get_department_hoursGet department hoursGETAgents: ask firstReturns the dealership's operating-hours sets (one per department or agent) with their opening and closing times and time zone. Read it before update_department_hours.
update_department_hoursUpdate department hoursPUTAgents: ask firstUpdates the opening and closing times of existing hours sets, identified by hours_id from get_department_hours. Changes the hours the AI gives customers. Answers with update counts and the new hours.

Json, 2 operations

upload_new_inventory_jsonUpload new inventory jsonPUTAgents: ask firstUploads a JSON file of new cars that REPLACES the dealership's whole new-car inventory; the answer counts the vehicles removed and added. DealerAI does not document the file's layout.
upload_used_inventory_jsonUpload used inventory jsonPUTAgents: ask firstUploads a JSON file of used cars that REPLACES the dealership's whole used-car inventory; the answer counts the vehicles removed and added. DealerAI does not document the file's layout.

Knowledge base, 2 operations

list_knowledge_base_entriesList knowledge base entriesGETAgents: ask firstLists the dealership's knowledge base entries (the facts the AI answers from), optionally filtered by entry ids, topics or text.
save_knowledge_base_entriesSave knowledge base entriesPUTAgents: ask firstCreates or updates knowledge base entries (pass info_id from list_knowledge_base_entries to update one). Changes what the AI tells customers; DealerAI retrains it when content changes.

Active, 1 operation

set_promotion_activeSet promotion activePATCHAgents: ask firstTurns a promotion on or off: active true lets the AI offer it, false stops it. Returns the updated promotion.

Adf, 1 operation

submit_adf_leadSubmit adf leadPOSTAgents: ask firstSends a lead to DealerAI as an ADF (Auto-Lead Data Format) XML document. Per DealerAI's guide, an inbound ADF lead starts the matching inbound sequence, whose AI then messages the customer. Answers 202 Accepted with no data.

Contact, 1 operation

upsert_contactUpsert contactPOSTAgents: ask firstCreates a contact, or updates the existing one matched by email or phone. Sends no message to the contact. DealerAI does not document the fields of the answer.

Dealership, 1 operation

get_dealershipGet dealershipGETAgents: ask firstReturns the dealership's profile and AI setup: name, region, links, phones and emails, per-agent appointment and lead settings, Reach, calculators, languages, Xtime. Integration keys and the Slack webhook are withheld.

Enrollment, 1 operation

enroll_contactEnroll contactPOSTAgents: ask firstEnrolls a person in a sequence (creating the contact when none matches): the sequence then MESSAGES them on its channels from next_message_date_time_utc on. Returns contact_id and enrollment_id.

Entry, 1 operation

delete_knowledge_base_entriesDelete knowledge base entriesDELETEAgents: ask firstDeletes knowledge base entries by info_id and/or by topic (descriptor). Cannot be undone; DealerAI retrains the AI when content changes. The answer lists the entries left.

Extract image, 1 operation

extract_promotion_from_imageExtract promotion from imagePOSTAgents: ask firstHas DealerAI's AI read a promotion image (by URL) and save it as a promotion, for dealers that publish offers as images. Returns the promotion, or no data when an existing one is left untouched.

Location, 1 operation

update_locationUpdate locationPUTAgents: ask firstUpdates the dealership's location (address, coordinates, map links). Leave out a field to keep its stored value; location_id is needed only when the dealership has several. Location data is retrained on any change.

Manufacturer, 1 operation

list_manufacturersList manufacturersGETAgents: ask firstLists the vehicle manufacturers (makes) known to DealerAI, paged with Skip and Take. Pass an id to list_manufacturer_models.

Model, 1 operation

list_manufacturer_modelsList manufacturer modelsGETAgents: ask firstLists a manufacturer's car models with their trims and specifications, optionally only those available in one region, paged with Skip and Take.

Replace, 1 operation

replace_knowledge_baseReplace knowledge basePUTAgents: ask firstReplaces the WHOLE knowledge base with the entries sent: creates, updates, reorders, and removes every entry left out. Read list_knowledge_base_entries first and send the full set back with your change.

Stop, 1 operation

stop_enrollmentStop enrollmentPOSTAgents: ask firstStops an enrollment: its scheduled messages are cancelled and the contact leaves the sequence (and can be enrolled again); lead extraction runs as for a manual unenroll. Repeating it returns the status.

Sync statutory, 1 operation

sync_statutory_holidaysSync statutory holidaysPOSTAgents: ask firstSyncs the upcoming statutory holidays into the dealership's holiday calendar (like the portal's Auto-fill), creating and removing entries. The answer counts both and says whether the AI was retrained.

Train llm, 1 operation

retrain_aiRetrain aiPOSTAgents: ask firstRetrains DealerAI's AI on one data set: KnowledgeBase (also staff and department contacts), Holidays, Locations or Services. Run it after changing that data so the assistant uses it.

Translate, 1 operation

translate_car_searchTranslate car searchPOSTAgents: ask firstTurns a natural-language car search into DealerAI's structured filter (attribute, comparator, value, nested arguments) without searching. DealerAI runs an AI model for it.

Verify, 1 operation

verify_credentialsVerify credentialsPOSTAgents: ask firstChecks the connected DealerAI API username and password and returns the authenticated dealer id (DealerAI does not document the layout of the answer). Changes nothing.

Version, 1 operation

get_versionGet versionGETAgents: ask firstReturns the version of the DealerAI API application. Answers without credentials.

Use DealerAI from Claude Code, Cursor or Codex

Connect DealerAI in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its DealerAI operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect DealerAI

Create your workspace
Sign up free. Nothing to install.
Connect DealerAI once
Add your DealerAI connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.