AI Product Detective

Retrieve progress or the complete assessment report for a scan owned by the connected user. Includes evidence, observed model recommendation counts, price discrepancies and suggested improvements. Product matching is conservative; inspect alternative product names and distinguish brand recommendations from confirmed product matches. API results do not reproduce consumer apps. Failed and unverified search tests are excluded from counts.

get_product_scan

List up to 20 unexpired assessments owned by the connected account and the remaining monthly assessment allowance with its renewal time. Use to find the user’s existing report or check assessment availability.

list_product_scans

Use when a merchant wants to assess whether AI models recommend their product or report its price and description accurately. Read one public HTTPS product page and save a private assessment draft for the user to confirm. Does not run model tests or use the assessment allowance. Assessments use model APIs, not consumer ChatGPT, Claude or Gemini app results. Do not use for ordinary shopper product recommendations.

prepare_product_scan

Start the user-confirmed draft assessment. Uses one included assessment and queues 21 model API tests. Requires an explicit request to start. Repeating for the same scan is safe and returns its existing state. Returns promptly; use get_product_scan to check completion.

start_product_scan